Tag: Budget Management

Claude-Powered PPC Automation: From Prompts to Systems

If Claude gives you a strong search-term analysis only after you paste the same instructions and CSV into a new chat, you have improved the task, not automated it. You still have to assemble the context, request the analysis, normalize the output, and move each approved change into Google Ads.

Claude-powered PPC automation becomes useful when you design those handoffs once. The practical system has three separate parts: decision logic, access to current campaign data, and controls over what the AI may change. Get those parts right and Claude can take recurring work off your desk without taking campaign authority away from you.

The three parts of a reliable Claude PPC system

Three connected modules represent campaign data access, AI decision logic, and human-controlled execution safeguards.

The model is only one layer of the system. A dependable workflow also needs a stable playbook and an explicit operating boundary.

System part	What it does	The question you must answer
Claude Skill	Encodes the task, decision rules, required inputs, exceptions, and output structure.	What should happen every time this PPC job runs?
Data and tools	Supply campaign context and, when authorized, provide a way to execute an approved action.	Which data may Claude read, and which operations may it call?
Workflow controls	Define scope, approval requirements, stop conditions, and records of proposed or completed changes.	What is Claude allowed to decide, recommend, and change?

A Claude Skill is a task-specific playbook, not a general preference about tone or behavior. It can tell Claude how to audit an account, evaluate search terms, generate ad assets, or compare budget opportunities. The instructions can be stored in a Markdown file, kept locally, or shared through a repository so the team uses the same method.

The main benefit is procedural consistency. Without a fixed contract, one run might return letter grades while another uses percentages or an unrelated numerical scale. That is more than a presentation problem. A person, spreadsheet, script, or approval workflow cannot reliably consume an output whose structure changes between runs.

A Skill should make the process predictable, but it should not pretend every PPC judgment is deterministic. Campaign evidence changes, and some cases will remain ambiguous. Your playbook therefore needs both decision rules and an explicit way to return insufficient evidence, conflicting signals, or required human review.

The data layer solves a different problem. A Skill can know how to evaluate a search query report while knowing nothing about the queries currently appearing in your account. A Model Context Protocol connection can bridge that gap: MCP can connect Skill logic to live data sources and account tools. That turns a static playbook into an operating workflow, but it also makes permissions and approval gates essential.

Build the first workflow around one recurring decision

Start with a bounded job rather than asking Claude to optimize an account. Search-term mining is a practical first candidate because you can define the input, inspect every recommendation, and test the logic without granting write access.

Define the job in one sentence. For example: review search terms from the requested 14-day window, identify waste and opportunity using the account’s approved criteria, and return proposed actions for review. The 14-day period is an input to this workflow, not a universal recommendation for every account.
Write down the judgment currently living in the operator’s head. Include the evidence Claude must consider, the conditions that support each recommendation, the exceptions that require escalation, and anything it must never infer from missing data.
Lock the output contract. Name every required field, its allowed values, and what a stopped run looks like. Do not let Claude invent a new scoring system or column set each time.
Convert the SOP into a Skill. A useful instruction is: Convert this SOP into a task-specific Claude Skill. Preserve the decision rules, define required inputs, return a fixed schema, stop when required fields are missing, and do not take write actions without approval.
Run the Skill against a known CSV before connecting an account. Confirm that it covers the intended records, follows the rubric, flags exceptions, and returns the exact structure your reviewer or downstream tool expects.
Connect live data in read-only mode. Compare the live run with the CSV-based process. Add write capabilities only after the connected workflow passes the same acceptance checks.

A useful output contract for this workflow can require:

The account, campaign, and reporting window included in the run.
A completion status that distinguishes a finished analysis from a stopped or incomplete run.
The item reviewed, the evidence used, and the applicable decision rule.
The proposed action and a concise reason for it.
An exception field for missing inputs, conflicting signals, or cases outside the Skill’s authority.
An authorization state such as proposal, approved, executed, or rejected.

The output contract is what turns a clever response into a component another person or system can trust. Claude should never quietly substitute a plausible answer when a required campaign field is unavailable. A stopped run with a precise error is safer and more useful than a polished recommendation built on incomplete context.

Put money-changing actions behind explicit gates

A human operator approves one proposed campaign change at a guarded barrier before it reaches an advertising budget.

Access and authority are not the same thing. An MCP-enabled tool may make an account change technically possible, but your workflow still decides whether Claude may propose it, prepare it, or execute it. That distinction matters whenever an action can change spend, targeting, messaging, or delivery.

Operating mode	Claude’s role	Human role
Manual-context assistant	Analyzes an uploaded report and returns structured recommendations.	Exports data, checks the result, and implements every change.
Connected analyst	Pulls permitted live data and prepares account-specific proposals.	Reviews and approves each proposed action before execution.
Controlled operator	Executes only approved action types within the defined scope and constraints.	Sets policy, handles exceptions, reviews logs, and can stop the workflow.

Most teams should move through these modes in order. Live read access removes manual report handling without immediately exposing the account to automated edits. Proposal-only operation then shows whether the logic behaves well under current conditions. Controlled execution comes last, after the team knows which exceptions appear in real runs.

Before enabling any write action, add these controls to the workflow:

Default-deny permissions. Claude may read or modify only the accounts, campaigns, objects, and action types explicitly included in scope.
Action-specific approval. Treat applying an existing extension, creating an ad experiment, changing a search-term response, and reallocating budget as separate permissions.
User-defined financial boundaries. A budget workflow must operate inside limits set by the account owner rather than deciding its own acceptable spend change.
Fail-closed behavior. Missing data, an invalid schema, an unavailable tool, or an out-of-scope request should stop the run instead of triggering a best guess.
A preview of the exact modification. The reviewer should see what object will change, its current state, the proposed state, and the reason before approving it.
An audit trail. Preserve the input scope, Skill version, findings, approval state, tool response, and execution result so a later reviewer can reconstruct what happened.
A conflict rule. Give each task one canonical Skill, because overlapping audit or optimization Skills can reintroduce the inconsistency the system was built to remove.
A recovery plan. Document how an executed change will be reversed when reversal is available. Keep irreversible or poorly understood actions manual.

Budget reallocation deserves the tightest gate because it moves money between campaigns. A recommendation can still be automated: Claude can compare the permitted data, explain the proposed shift, and prepare the action. Execution should remain subject to the account owner’s constraints and approval until the workflow has demonstrated reliable behavior in proposal-only mode.

Use acceptance checks rather than impressions when deciding whether a workflow is ready. The run should always return the required fields, stop on missing inputs, stay inside its declared scope, expose the evidence behind each proposal, and show the planned modification before execution. If any of those checks fail, improve the Skill or connection before expanding its authority.

Choose PPC tasks by controllability, not novelty

The best first automation is not necessarily the task consuming the largest budget or producing the most visible output. It is the task whose rules can be written clearly, whose evidence can be inspected, and whose mistakes can be contained.

PPC workflow	What the Skill should standardize	First safe deployment	Expanded deployment
Search-term mining	The evaluation rubric, required evidence, exception handling, and recommendation format.	Analyze an uploaded report and return proposals for review.	Pull live search-term data and implement only separately approved actions.
Ad copy generation	How landing-page information, keywords, user intent, and value propositions become proposed ad assets.	Generate structured drafts for human review.	Identify underperforming ads, prepare alternatives, and create an approved experiment.
Account auditing	The checklist, severity logic, supporting evidence, and distinction between findings and remedies.	Return a consistent audit with no account changes.	Use live account data and apply permitted remedies, such as attaching an existing extension where appropriate.
Budget reallocation	The comparison method, constraints, explanation, and escalation conditions.	Produce proposed reallocations with no write access.	Execute approved shifts inside account-owner limits and record every result.

These four workflows can all progress from manual data handling to connected execution, but they should not receive the same authority by default. Search-term analysis, ad generation, account auditing, and budget reallocation involve different consequences and therefore need different approval paths.

Score a candidate workflow against five practical questions before building it:

Does the task recur often enough that removing handoffs will matter?
Can an experienced operator state the decision rules without relying on unexplained instinct?
Are the required inputs available in a stable, inspectable form?
Can a reviewer verify the recommendation before the account changes?
Can the impact of an error be contained to a narrow scope?

If the answers are weak, connecting more tools will not improve the workflow. Clarify the SOP first. Automation magnifies whatever is encoded: good judgment becomes repeatable, while an ambiguous process becomes ambiguous at greater speed.

For a first deployment, we would favor a proposal-only search-term or account-audit workflow. Both make it easy to compare Claude’s output with an existing human process. Ad experiments can follow once asset review is defined. Budget execution belongs later because its consequences reach spend directly.

Frequently asked questions

What is Claude-powered PPC automation?

It is a workflow in which a Claude Skill applies a repeatable PPC playbook, data connections supply the required campaign context, and explicit permissions determine whether Claude analyzes, proposes, or executes an action. A chat response alone is assistance; automation also handles the recurring context and handoffs.

Do you need MCP to use a Claude Skill for PPC?

No. You can run a Skill against a manually uploaded CSV and implement its recommendations yourself. MCP becomes relevant when you want Claude to retrieve live data or use connected account tools. Start with manual or read-only data if the Skill’s decision logic has not yet been validated.

Which PPC workflow should you automate first?

Choose a recurring workflow with written rules, inspectable inputs, a fixed output, and limited consequences when something goes wrong. Search-term mining or a checklist-based audit is usually easier to validate than autonomous budget reallocation. Keep the first version proposal-only so you can judge the logic before granting execution authority.

How do you prevent inconsistent Claude outputs?

Use one canonical Skill for the task, define required fields and allowed values, state how exceptions must be returned, and stop the run when required data is missing. Remove or narrow competing Skills that could handle the same request. Test structural consistency before connecting the output to another tool.

Take the next recurring search-term review or account audit and write down its rubric, output contract, and stop conditions. Test that process on a CSV, connect live data in read-only mode, and grant write access only after the workflow passes explicit acceptance checks. That sequence turns Claude from another prompt window into a PPC system you can supervise.

References

Search Engine Land – Transforming PPC with Claude Skills for Automation Success

April 14, 2026

Google Ads MCC Security: A Practical Recovery Playbook

You are locked out of a Google Ads Manager Account, unfamiliar administrators are appearing, or client billing has started changing without approval. Treat that as an active identity, advertising, and financial incident. Your first job is not to restore the MCC dashboard. It is to stop the compromised manager from reaching more client accounts.

The recovery order matters. Contain accounts through access you still trust, secure the Google identities behind that access, escalate every affected Customer ID, and only then rebuild the manager hierarchy. This playbook gives you a practical sequence for doing that without mistaking a restored login for a clean account.

Treat the manager account as hostile until you contain it

A Google Ads Manager Account, still commonly called an MCC, concentrates access. That makes it operationally convenient and potentially dangerous: one compromised administrator can expose multiple client accounts, manager relationships, campaigns, and billing arrangements.

Once you see a credible takeover signal, stop using the affected MCC as your control center. An attacker with administrative access may be able to remove your users, alter allowed-domain settings, create another manager account with a familiar company name, issue invitations, change payment arrangements, and launch unauthorized campaigns. Work from client-owned accounts and clean identities wherever possible.

Open an incident record outside the affected account. Record the detection time, every known Customer ID, the manager hierarchy you expected, suspicious email addresses, unauthorized campaigns, billing changes, and every action your team takes. Assign one person to maintain the timeline so simultaneous recovery work does not create conflicting instructions.
Identify access you can still trust. Contact each client’s known account owner through a previously established channel. Ask an existing client administrator to sign in directly, confirm that their own Google identity is secure, and inspect the account without relying on an invitation sent during the incident.
Disconnect exposed client accounts from the compromised MCC. A client administrator with retained access can remove the manager relationship and preserve an independent route into the account. Coordinate this with the client because disconnecting an agency manager can interrupt normal management workflows, but leaving a hostile manager attached preserves the attacker’s reach.
Escalate every affected account to Google. Contact your established Google representative if you have one, and use Google’s compromised-account process. Submit an Account Takeover Form for each affected client account and for the MCC itself. Do not assume a single manager-level report automatically creates cases for every linked Customer ID.
Control advertising exposure. Through clean client access, inspect recently created or materially changed campaigns, budgets, ads, and destination URLs. Pause clearly unauthorized activity when you have confirmed that doing so will not stop legitimate campaigns. Keep a record of what you paused and why.
Bring the authorized billing owner into the incident. Review the payment manager, payment methods, failed or pending charges, and any unfamiliar billing profile changes. Ask the bank or card issuer about suspicious attempts. Do not indiscriminately delete payment information or replace billing ownership without documenting the existing state; that can disrupt legitimate campaigns and make reconciliation harder.

The potential blast radius is not theoretical. In one documented MCC takeover, administrators were removed, the allowed domains were changed to admit Gmail addresses, more than a dozen people were invited to a newly created manager account, payment arrangements were altered, and unauthorized campaigns appeared. Attempted fraudulent charges reached half a million on some accounts. Control was restored within eight hours and the direct loss was limited to $100, but that outcome is an incident example, not a recovery-time or loss benchmark.

If you cannot reach a clean administrator, do not create a new relationship through an identity that may also be compromised. Preserve the Customer ID and other evidence, continue the Google escalation, and involve a cybersecurity professional when the attacker remains active across multiple email accounts or devices.

Recover each client account in a controlled order

Gloved hands isolate account modules and secure identity tokens in a staged incident-response workspace.

Containment removes or limits the attacker’s path. Recovery proves that each layer is clean. Getting back into the MCC does not establish that its administrators, manager links, payment manager, or client campaigns are safe. Reconnecting every client immediately can restore broad access before you know whether the underlying identity breach has been removed.

Create a recovery worksheet from records outside the compromised hierarchy: contracts, client contact lists, prior invoices, Customer ID inventories, and configuration backups. Track every client separately. At minimum, include the Customer ID, trusted client administrator, expected manager relationship, takeover-case status, billing owner, suspicious changes, cleanup owner, and approval to reconnect.

Recovery layer	What to verify	Condition before sign-off
Google identity	Email security, passwords, active sessions, recovery methods, and 2FA enrollment for every retained user	Only verified people control the identities that will receive Ads access
Users and manager links	Administrators, invitations, allowed domains, linked managers, and any similarly named MCC	Every user and manager relationship has a documented business owner
Billing	Payment manager, payment methods, billing profiles, failed attempts, pending charges, and client authorization	The client or authorized finance owner confirms the intended arrangement
Campaign configuration	New campaigns, budgets, ads, destinations, schedules, and other changes made during the incident window	Unauthorized changes are reversed or paused and legitimate changes are preserved

Use Google Ads change history to build the account-side timeline. Start slightly before the first visible symptom and follow the sequence forward. Look for user removals, invitations, domain-setting changes, new manager relationships, billing modifications, campaign creation, budget changes, and cleanup attempts. Detailed timestamps can help you distinguish the attacker’s actions from the emergency changes made by your own team.

Record the earliest suspicious Ads change and the identity associated with it.
Match later changes to the account, campaign, billing, or manager layer they affected.
Mark your own emergency actions so they are not mistaken for attacker activity.
Compare the final configuration with a known-good export or Google Ads Editor backup.
Keep unresolved items open instead of treating restored access as proof that they are harmless.

Change history is valuable, but it is not a complete identity-forensics record. It can show what changed in Google Ads and when; it may not prove how an employee mailbox was first compromised. Pair it with the security activity available for the affected Google identities and with your internal email, device, and access records.

Reconnect a client only after its trusted administrator approves the user list, manager relationship, billing state, and campaign configuration. Use the original verified Customer IDs rather than accepting a link merely because the manager account has your agency’s name. A copycat MCC can look convincing while remaining fully controlled by an attacker.

Investigate the identity breach even when 2FA was enabled

Two-factor authentication is an important control, but its presence does not prove that an identity is clean. If an attacker has maintained access to an employee’s email account, recovery settings, or approved device, that attacker may be able to establish an authentication path that looks legitimate. Resetting only the Google Ads password can leave that path intact.

In the documented takeover, the attackers tried multiple employee identities before succeeding through a junior employee’s email. That email had apparently been compromised for months, and the attackers had configured their own 2FA before taking over the MCC. Phishing or a compromised password was considered a likely initial route, but the exact entry method was not established. The lesson is precise: investigate the user’s wider Google identity and device sessions, not just the Ads permission that was abused.

Secure the mailbox first. Change any compromised or reused password, review recovery options, remove unfamiliar access, and revoke active sessions. A unique password for every service limits the chance that credentials exposed elsewhere can be reused against a Google identity.
Rebuild 2FA enrollment. Remove authentication methods you cannot attribute to the user and enroll a dedicated method under a controlled process. Do not rely solely on device approval notifications that a user can accept reflexively or that an attacker-controlled device may receive.
Review every person with MCC access. Check administrators as well as standard users. A low-privilege employee identity can still become an entry point if it has more Google Ads permissions than the role requires.
Revoke old sessions and devices. A password change is not the same as a complete session reset. End existing sessions as part of the recovery so a previously authenticated attacker is not silently left connected.
Restore the minimum required Ads role. Give the returning user only the access needed for current work. Do not restore administrative access merely because the person held it before the incident.

Apply the same skepticism to invitations. Tell clients that unexpected Google Ads access or manager-link requests must be verified with a known agency contact through a separate channel. They should not confirm legitimacy by replying to the invitation email or by trusting the manager’s display name. In the takeover described above, clients avoided a larger problem by ignoring invitations from the fraudulent manager.

If suspicious access reappears after passwords, sessions, and 2FA have been reset, stop cycling credentials without a broader investigation. Persistent access can indicate that another mailbox, recovery route, device, or administrator is still compromised. That is the point to involve an identity-security or incident-response specialist.

Build an MCC that can fail without taking clients with it

A secure cutaway complex separates client account chambers, access checkpoints, payment vaults, and backup routes into isolated compartments.

The strongest MCC design does not depend on preventing every credential attack. It limits what one compromised identity can reach and preserves a clean way back into every client account. That requires changes to ownership, permissions, authentication, billing, backups, and escalation procedures.

Keep independent client administrators. Every client should retain access to its own account through an identity the client controls. This is good account governance and a recovery mechanism: the client can inspect activity or disconnect a compromised manager without waiting for the agency’s MCC to be restored.
Use least privilege for agency users. Reserve administrative access for people who actually manage users, manager relationships, security settings, or billing. Campaign operators should receive the lowest role that supports their work. Reassess permissions when responsibilities change instead of allowing access to accumulate.
Remove stale surface area. Unlink obsolete client accounts and unused MCCs, remove former employees and contractors, close unnecessary invitations, and review allowed domains. A dormant relationship can remain useful to an attacker even when nobody on your team remembers it exists.
Run recurring access recertification. Ask a named owner to affirm every user, manager link, and administrative role. Do not treat a spreadsheet export as a completed review; each entry needs a business reason and an accountable owner.
Use unique passwords and controlled authentication. Each person should have an individual identity rather than a shared login. Enroll 2FA deliberately, favor dedicated authenticators over casual notification approvals, and include session revocation in suspected-compromise procedures.
Enable multi-party approval where available. Google’s multi-party approval feature requires another administrator to confirm certain major changes. It reduces the chance that one compromised administrator can complete a sensitive action alone. The second approver must use a separately secured identity or the control becomes ceremonial.
Keep recoverable campaign configurations. Export regular backups with Google Ads Editor, and refresh them after approved material changes. Store them somewhere the compromised MCC cannot alter. A backup will not restore identity or billing ownership, but it gives you a known configuration against which to identify and reverse campaign changes.
Design billing escalation before an incident. Document who can change the payment manager, who speaks for each client, and who contacts the bank or card issuer. Credit or invoice arrangements helped financial institutions flag irregular transactions in the documented takeover, but no payment method guarantees that fraud will be stopped. Your finance owner still needs a rapid review process.
Maintain human escalation routes. Keep current contact details for Google representatives, client administrators, finance owners, internal security staff, and agency peers who can help establish the scope. Store this list outside Google Ads so it remains available when the MCC does not.

Test the design by assuming the MCC is unavailable. Can you name every linked Customer ID? Can each client reach its account independently? Can your team find a known-good campaign export? Does everyone know who is allowed to request a manager link and how that request is verified? Any answer that exists only inside the MCC is a dependency worth fixing.

Key takeaways for Google Ads MCC security and recovery

Treat a suspected MCC takeover as an identity, advertising, and billing incident, not merely a lost-password problem.
Use verified client-owned admin access to disconnect exposed accounts and reduce the compromised manager’s reach.
Submit a takeover case for every affected Customer ID, including the manager account, and keep one external incident timeline.
Validate identities and sessions before restoring Ads permissions; 2FA does not help if the attacker enrolled or controls the second factor.
Reconcile users, manager links, billing, and campaign changes before reconnecting a client to the recovered MCC.
Reduce future impact with independent client admins, least privilege, access recertification, multi-party approval, and Google Ads Editor backups.

Set up one recovery drill before you need it. Export the current client and manager inventory, confirm an independent administrator for every client, save a known-good configuration, and put the escalation contacts where the team can reach them without the MCC. The useful standard is simple: if the manager account disappeared tonight, you could still identify, contact, contain, and recover every client account.

References

CrushPress.AI — Rescuing Your Google Ads MCC from Hackers

April 14, 2026

A Practical PPC Strategy for Using AI Without Losing Control
Your PPC account may already use automated bidding, AI-assisted targeting, and generative tools, yet still leave you unsure whether the system is making good decisions. That uncertainty usually isn’t a reason to abandon AI. It is a reason to tighten the operating system around it.

A dependable AI-assisted PPC strategy has a clear order: verify the business data, simplify the account around meaningful decisions, constrain automation where failure would be expensive, and turn every recommendation into a test. Follow that order and AI becomes easier to trust because its work remains visible, measurable, and reversible.

Start by proving that your ROAS means what you think

Your first AI decision isn’t which model, campaign type, or bidding strategy to use. It is whether the conversion value entering the system represents the business result you intend to optimize.

ROAS is calculated by dividing attributed conversion value by advertising cost. The calculation is simple, but the inputs can be misleading. A dashboard may produce a precise ratio even when its currencies, conversion actions, or imported revenue values are inconsistent.

Currency errors are particularly dangerous because they can affect reporting without producing an obvious technical failure. In one paid-media account, a mismatch between an Australian billing setup and reporting in GBP distorted conversion values so severely that CRM reconciliation showed actual performance was twice the reported level. An optimization decision made from the advertising dashboard alone would have started from the wrong diagnosis.

Before changing bids, budgets, or account structure, audit the measurement chain:
- Confirm the account billing currency, conversion-value currency, and reporting currency. Document every intentional conversion between them.
- Compare advertising-platform revenue with CRM, order-management, or finance data over equivalent reporting periods.
- Verify which conversion actions are included in bidding. Remove duplicate or secondary actions from the primary optimization signal unless they represent genuine incremental value.
- Check whether cancellations, refunds, offline sales, and qualified leads are handled consistently.
- Record the attribution view and normal conversion delay so that the team does not compare numbers built on different rules.
- Name the system that decides the final business outcome. The ad platform may guide bidding while the CRM remains the authority for lead quality or realized revenue.
Treat disagreement between systems as a diagnostic signal, not an inconvenience to average away. If the platform and CRM both decline, the performance problem may be real. If platform revenue falls while CRM revenue remains stable, investigate tracking, attribution, currency, and reporting logic before rebuilding campaigns. If platform ROAS rises while qualified revenue stays flat, the bidding system may be optimizing toward a convenient but weak proxy.

This audit protects more than reporting accuracy. Automated bidding learns from the values you send it. A corrupt value is therefore both a measurement problem and an instruction to spend money in the wrong places.

Build an account structure AI can learn from

Many legacy PPC structures were designed when control meant separating almost every keyword, product, market, or match type. That granularity made sense when practitioners performed more decisions manually. It can work against automated systems when it fragments related data and creates thousands of campaign-level boundaries.

A travel account built around thousands of legacy campaigns exposed the conflict between an old granular structure and modern AI bidding. The lesson is not that every account should be aggressively consolidated. It is that every separation should still earn its place.

Keep campaigns separate when the boundary changes a real decision, such as:
- A distinct business objective or conversion action.
- A materially different margin, customer value, or acceptable acquisition cost.
- A budget that must be protected or controlled independently.
- A geographic, language, regulatory, inventory, or landing-page difference that changes eligibility or performance.
- A brand-protection requirement or an exclusion that cannot safely be shared.
Consider combining structures when the only distinction is an inherited naming convention, a reporting preference that can be handled with labels, or a keyword taxonomy that does not change bidding economics. The goal is not the fewest possible campaigns. It is the fewest boundaries needed to express genuine business constraints.

Restructure in stages rather than replacing the account in one irreversible move:
1. Map every existing campaign to its objective, budget owner, conversion signal, audience, destination, and economic target.
2. Mark the boundaries that affect business decisions and the ones that exist only because of account history.
3. Capture a clean performance baseline and annotate known tracking or seasonal issues.
4. Migrate a representative, lower-risk portion first. Check query routing, budgets, conversion recording, and lead or revenue quality.
5. Expand only after the new structure behaves as intended. Retain a documented rollback path while the change is being evaluated.
Timing matters as much as architecture. A peak trading period is a poor moment for a sweeping rebuild, but indefinite postponement creates its own risk. One delayed restructuring effort had to be accelerated after performance weakened in January, creating the pressure the delay was meant to avoid. Choose a lower-risk implementation window with enough runway to validate the new structure before the next commercially critical period.

Put guardrails around automated bidding

Using automation does not require giving the platform unlimited freedom. Your job is to define the objective, provide trustworthy signals, decide which decisions the system may make, and set boundaries around outcomes the business cannot tolerate.

Useful controls can include campaign budgets, portfolio boundaries, eligible locations, audience exclusions, conversion-action selection, inventory rules, and bid limits where the chosen platform and strategy support them. Pick the control that addresses the observed failure mode. Do not add constraints merely to make an automated campaign feel more manual.

A max CPC cap applied within portfolio bidding once reduced click costs without damaging performance. That is evidence that a well-chosen boundary can improve an automated system, not proof that every account needs the same cap. A cap set below the price of useful auctions can suppress traffic, conversion volume, and access to high-value prospects.

Use a guardrail protocol whenever you intervene:
1. Name the failure precisely. Runaway CPC, weak lead quality, overspending in one segment, and volatile total spend are different problems.
2. Capture the baseline. Record CPC, click volume, conversion volume, attributed value, and the CRM outcome that matters to the business.
3. Change one meaningful control. If you alter the cap, budget, targeting, and conversion setup together, you will not know which change produced the result.
4. Allow for the normal conversion cycle. A constraint can look efficient immediately because it reduced traffic, while its effect on qualified revenue appears later.
5. Judge the business result. Lower CPC is not a win if profitable volume, lead quality, or realized revenue also falls.
6. Keep the decision reversible. Define in advance which outcome means keep, loosen, or remove the constraint.
This is the practical middle ground between blind automation and constant manual interference. The algorithm keeps enough freedom to respond to auctions, while you retain control over the economics and risk.

Prompt generative AI like an analyst with a proper brief

Generative AI can help categorize search terms, surface anomalies, draft testing hypotheses, explain performance patterns, and prepare questions for deeper analysis. It cannot rescue an underspecified request. Detailed context about the goal and audience produces more useful output than a vague prompt.

A request such as analyze this campaign gives the model no reliable definition of success. It does not know whether you care about revenue, qualified leads, margin, new customers, market coverage, or budget stability. It also does not know which fields are facts, which are calculated metrics, or which constraints it must respect.

Build PPC prompts from seven parts:
1. Context: Describe the business model, campaign type, funnel stage, audience, and decision you face.
2. Objective: State the business outcome and the advertising metric being used as its proxy.
3. Data definitions: Explain the reporting period, currency, attribution view, conversion delay, and meaning of each important field.
4. Evidence: Supply only the relevant account, CRM, and historical information. Label missing or unreliable fields.
5. Constraints: Include budget limits, brand rules, geographic boundaries, minimum volume requirements, and changes that are off limits.
6. Task: Ask for a specific deliverable, such as ranked hypotheses, an anomaly check, or a test plan.
7. Output rules: Require the model to separate observations from inferences, identify missing evidence, and state what would disprove each recommendation.
A reusable prompt frame can be short: Review the supplied PPC and CRM data to explain the change in qualified revenue. Use the definitions and constraints below. Separate measured facts, plausible causes, and unsupported possibilities. Rank the hypotheses by evidence strength. For each one, give the confirming evidence, conflicting evidence, next check, and safest reversible test. Mark missing information as unknown rather than filling the gap.

That last instruction matters. Fluent output can conceal uncertainty. Asking for competing explanations and disconfirming evidence makes it easier to spot a recommendation that merely sounds plausible.

Protect client and customer data as you work. Remove customer-level identifiers, avoid pasting credentials or confidential commercial details into unapproved tools, and follow the data-use rules that apply to your organization. AI assistance does not change your responsibility for access control or final decisions.

Turn every change into a controlled learning loop

A test-and-learn culture is not permission to make a stream of undocumented changes. It is a discipline for converting uncertainty into evidence without putting the whole account at risk.

For every material test, create a decision record containing:
- The problem being addressed and the evidence that it exists.
- The hypothesis linking the proposed change to the expected outcome.
- The primary business metric and the secondary indicators that guard against a hollow win.
- The account segment affected and what remains unchanged for comparison.
- The expected conversion delay and the condition for making a decision.
- The owner, approval, annotation, and rollback procedure.
Match the evaluation cadence to the account’s conversion volume and sales cycle. A low-volume campaign should not be judged with the same rhythm as a high-volume retail account, and a lead-generation campaign should not be declared successful before downstream quality becomes visible.

When performance falls, resist the urge to stack speculative fixes. Validate tracking and currency first. Review recent account and site changes. Locate whether the decline is concentrated by campaign, query class, audience, location, device, or conversion action. Reconcile platform outcomes with CRM results. Then decide whether the evidence supports a rollback, a targeted constraint, or continued observation.

Small mistakes still happen: an incorrect report is sent, a setting is misunderstood, or a change produces an unexpected effect. Fast acknowledgement protects the account better than defensive explanation. Correct the immediate problem, document the cause, add the missing check, and return the team’s attention to the business outcome.

Key takeaways
- Reconcile platform reporting with CRM or realized revenue before treating ROAS as an optimization signal.
- Separate campaigns for real business constraints, not inherited naming or reporting habits.
- Use automation guardrails to address a defined failure mode, and evaluate their effect on profitable volume rather than CPC alone.
- Give generative AI the objective, definitions, evidence, constraints, and uncertainty rules an analyst would need.
- Record each material change as a reversible test with a hypothesis, decision condition, and rollback path.
Your next move should be small and diagnostic. Before launching another bid-strategy change, reconcile one important revenue view from ad click to CRM outcome. That check will tell you whether the account needs better automation, better structure, or simply better data.

References
- Search Engine Land — Maddie Lightening speaks on misreported ROAS, account structure chaos and AI mistakes
April 10, 2026
How to Audit Google Ads Data and Cut Spend Waste Safely
Your Google Ads account can report a better return while the underlying business gets less efficient. That happens when conversions are duplicated, low-value actions are treated as primary goals, delayed sales are missing, or automated bidding receives values that do not match real revenue.

So do not begin an efficiency audit by lowering bids. Use this order: validate the conversion signal, classify waste, protect proven demand, choose automation that fits the available data, and then check whether product data is steering Shopping spend correctly.

Treat conversion tracking as a bidding input, not a reporting detail

Automated bidding does not know which outcomes matter to your business. It knows which conversion actions and values you send. If a page view, unqualified lead, duplicate purchase, or inflated order value is marked as a primary outcome, the system can optimize successfully toward the wrong result.

Start by writing a plain-language definition for every primary conversion. A purchase conversion should represent a completed order, not a checkout visit. A qualified-lead conversion should represent the stage named in its label, not every form submission. If revenue arrives after the initial lead, keep the early event for diagnosis but base your main performance decision on the deepest reliably measured outcome available.
- Confirm the event: Identify exactly what user or business action causes the conversion to fire.
- Confirm the count: Check whether one business outcome can create multiple ad conversions. Repeat purchases may be valid; repeated firing for one order is not.
- Confirm the value: Reconcile conversion values and currency with the system that records actual orders, revenue, or accepted leads.
- Confirm the role: Separate primary actions used for bidding from secondary observations used for diagnosis.
- Confirm the delay: Compare results only after the normal lag between an ad interaction and the recorded business outcome has had time to mature.
Google’s consolidated enhanced-conversions system makes matching easier, but it does not replace this validation. Under the June 2026 consolidation, user-provided data can arrive through website tags, Data Manager, and API connections at the same time. You no longer have to choose a single implementation method for enhanced conversions for web or leads.

That broader intake can recover conversions that would otherwise be harder to match. It cannot correct an event that fires twice, turn an unqualified lead into revenue, or repair an incorrect order value. Think of enhanced conversions as a matching layer around a conversion definition that must already be sound.

A practical validation sequence
1. Choose one high-spend campaign and list the primary conversion actions affecting its bidding.
2. Trigger each action through a controlled test and verify that the expected event arrives once with the correct label and value.
3. Reconcile a complete period of platform conversions against the corresponding records in your order, CRM, or lead-management system.
4. Investigate missing outcomes, duplicate outcomes, unexplained value differences, and changes in the normal reporting delay.
5. Resolve the discrepancy before changing a bid target or using the platform’s reported return to move budget.
Existing enhanced-conversions users generally do not need to enable the consolidated feature again if the required customer-data terms have already been accepted. New setups can enable it under Goals, then Settings, under Customer data use; it can also be controlled for individual conversion actions.

User-provided data still creates privacy and compliance obligations, even when it is hashed or transmitted through an approved integration. Do not enable another input merely because the switch is available. Confirm the applicable customer-data and data-processing terms, your consent or other lawful basis, your privacy disclosures, and the fields your implementation is permitted to send. Involve your privacy or legal owner if that authority is unclear.

Separate obvious waste from performance that needs more evidence

A zero-conversion row is not automatically waste. It may be new, low volume, affected by reporting delay, or part of a longer path to purchase. Cutting every row at zero conversions selects against campaigns before they have had a fair opportunity to produce an outcome.

A better audit divides questionable spend into three classes:
- Structural waste: The traffic cannot produce the intended outcome. Examples include an irrelevant search term, an unavailable product, or a destination that does not support the advertised action. Act as soon as you verify the mismatch; waiting for more conversions will not make the traffic relevant.
- Performance waste: The traffic could convert, but it has accumulated enough impressions, clicks, spend, and mature outcomes to miss the account’s CPA or ROAS requirement. This class needs sufficient data before you pause or constrain it.
- Measurement uncertainty: Spend looks weak because conversions, values, or delays cannot be trusted. Repair measurement before making a budget decision unless the traffic is also structurally irrelevant.
A useful working hypothesis is that 20% to 30% of spend may underperform in an audited account. That is an audit prompt, not a universal benchmark and certainly not a quota to cut. If your analysis identifies only 8% of defensible waste, removing 20% would damage productive activity. If it identifies more, preserving the budget because it fits the plan would be equally hard to justify.

Build your review at the lowest level where you can take a meaningful action. Search-term data can reveal irrelevant queries hidden by campaign averages. Product-level data can reveal items consuming spend while generating no conversions or falling well below the required return. Campaign totals alone can allow a few strong components to conceal a long tail of loss.
1. Choose an evaluation period that includes the normal conversion lag and enough activity to judge the unit fairly.
2. Review search terms, products, and other actionable segments using impressions, clicks, spend, conversions, conversion value, CPA, and ROAS.
3. Mark definite mismatches separately from low-performing but plausible traffic.
4. For each performance outlier, inspect the query, product availability, feed information, landing-page path, conversion signal, and offer before assigning the cause to bidding.
5. Apply the narrowest corrective action: add an exclusion for irrelevant demand, repair the destination or feed, constrain a proven outlier, or pause a segment whose economics no longer work.
6. Record what changed, the reason, the decision period, and the metric that will determine whether the intervention worked.
Use CPA and ROAS for different questions. CPA is cost divided by conversions and works only when the counted outcomes are sufficiently comparable. ROAS is conversion value divided by cost and works only when the values are complete and economically meaningful. A strong reported ROAS can still be unattractive if revenue values omit cancellations, returns, fulfillment costs, or other business constraints, so reconcile the platform result with the financial view used to run the business.

Reallocate budget instead of cutting every campaign evenly

An across-the-board reduction feels neutral, but it removes money from proven demand and waste at the same rate. That can preserve the account’s weakest activity while forcing high-intent campaigns to stop serving earlier.

Protect lower-funnel activity that has trustworthy measurement, sufficient volume, and a return that meets the business requirement. Move money away from confirmed structural waste first, then from mature performance outliers. Keep uncertain activity in a clearly bounded diagnosis or testing budget so it cannot consume funds without an explicit decision date.
- Protected budget: Proven, high-intent activity meeting its business target with reliable tracking.
- Repair budget: Valuable demand whose feed, landing page, creative, or measurement problem has a credible fix.
- Test budget: New queries, products, audiences, or creative variations with a stated hypothesis and success criterion.
- Exit budget: Irrelevant demand and mature segments that remain outside acceptable economics after measurement problems are ruled out.
Do not let platform ROAS become the only judge. Compare it with actual revenue or qualified outcomes from the business system and with the combined effect of your channels. That blended view matters because lower-funnel campaigns can capture demand created elsewhere, while upper-funnel activity may look weak when judged only by the final recorded click. The answer is not to protect every awareness campaign; it is to give each stage a measurement question appropriate to its job.

Ask two separate questions during every reallocation. First, should this activity exist at all? Second, how much budget has it earned? Combining those questions creates bad choices: a useful campaign may receive too much money simply because it belongs in the plan, while an irrelevant segment may survive because its budget is small.

Match bidding and creative decisions to the signal you actually have

A bid strategy cannot compensate for a weak objective. Select it only after you know which conversion signal is reliable and what the business is trying to control.
- Maximize Clicks: Use it when acquiring traffic is genuinely the immediate goal or when a dependable conversion signal is not yet available. Do not evaluate it as though it were instructed to maximize sales.
- Target CPA: Use it when the primary conversions are reasonably comparable in value and the account can supply trustworthy conversion data. A lead target is useful only if the counted leads correspond to the quality level the business can afford.
- Target ROAS: Use it when conversion values vary and those values accurately represent the outcomes you want the system to favor. Bad values turn a revenue-aware strategy into an amplifier of accounting errors.
Automation needs boundaries as well as data. Keep exclusions current, prevent invalid products and irrelevant queries from competing for budget, and avoid changing targets merely to make the interface report a preferred status. If a target conflicts with the economics of the business, the target is wrong even when the campaign reaches it.

Creative is another control surface, not decoration. Automated campaigns need meaningful variations to learn which message, format, and offer fit different opportunities. Maintain a queue of distinct assets rather than superficial rewrites of the same claim. Review each variation after adequate exposure, retire clearly weak assets, and preserve the message differences so the next test answers a new question.

Human review remains necessary because the platform can optimize the target it receives without knowing whether that target reflects margin, lead quality, inventory constraints, or business priorities. Use automation to process the signal; keep responsibility for defining and auditing the signal with your team.

For Shopping campaigns, product data is spend control

Shopping efficiency begins before the auction. Titles, product identifiers, availability, inventory, promotions, and other feed attributes determine what can serve and how the system understands the offer. A bid adjustment is the wrong fix when the product data itself is incomplete, stale, or mapped incorrectly.

Google set April 22, 2026 as the start of Merchant API support in Google Ads Scripts and August 18, 2026 as the retirement date for the Content API for Shopping. The Merchant API transition is therefore both a continuity requirement and an opportunity to improve how product-data problems are detected.

The Merchant API uses modular sub-APIs and expands control over supplemental product data, local and regional inventory, promotions, product and store reviews, and notifications. Google Product Studio also introduces generative-AI capabilities. Treat those enhancements as optional improvements after the functional migration is correct; generated content does not compensate for missing inventory or a broken product mapping.
1. Inventory every dependency: Find scripts, scheduled jobs, feed tools, supplemental inputs, inventory updates, promotions, reviews, and alerts that still rely on the Content API.
2. Map each function: Identify the relevant Merchant API module and the credentials, permissions, fields, and error handling needed by that function.
3. Enable the Advanced API: Update Google Ads Scripts that require Merchant API access and remove assumptions tied only to the legacy response structure.
4. Validate in parallel: While both paths are available, compare product identifiers, item counts, availability, inventory, promotions, and reported errors rather than assuming a successful request means equivalent data.
5. Test failure handling: Confirm that authentication errors, rejected products, delayed inventory updates, and other exceptions produce an alert that someone owns.
6. Cut over deliberately: Retire the legacy dependency only after the new path has completed its scheduled runs and the resulting catalog state matches the expected business state.
The Notifications API can make product issues visible sooner, but an alert has value only when it identifies the affected item, the severity, and the person or workflow responsible for the response. Route urgent availability or rejection problems differently from informational feed changes.

Key takeaways
- Reconcile primary conversion counts and values with the business system before changing bids or budgets.
- Use enhanced conversions to improve matching, not to repair duplicate events, weak conversion definitions, or incorrect values.
- Remove structural waste immediately, but require mature data before classifying plausible traffic as a performance failure.
- Protect proven lower-funnel demand, isolate tests, and move budget from confirmed waste instead of cutting every campaign equally.
- Choose Target CPA, Target ROAS, or Maximize Clicks according to the quality of the available signal and the outcome each strategy is actually designed to pursue.
- For Shopping campaigns, complete and validate the Merchant API migration because feed integrity directly affects where spend can go.
Open one high-spend campaign and reconcile its primary conversion count and value over a fully matured period. If the numbers match your business records, audit its search terms or products for structural and performance waste. If they do not match, fix the signal first. Every later optimization depends on that distinction.

References
April 10, 2026

How to Build a Conversion-Focused PPC Strategy for Revenue

Your PPC dashboard says conversions are up. Revenue, order value, or sales quality says otherwise. That gap usually means the account is optimizing for the easiest recorded action, not the outcome your business actually needs.

A conversion-focused PPC strategy fixes the problem in a specific order: define the valuable outcome, improve the signals sent to the platform, separate different kinds of intent, and test changes against business value. Automation can then help you pursue the right result instead of efficiently producing the wrong one.

Start with the conversion signal you actually want

A marketer redirects a conversion signal from a large pile of interaction tokens toward completed orders, payment confirmation, and a qualified customer.

A conversion is whatever your tracking setup labels as a conversion. It isn’t automatically a sale, a qualified lead, or a profitable customer.

This distinction matters because automated bidding learns from the outcomes you feed it. If a content download, an unqualified form submission, a valuable phone call, and a completed purchase all look equivalent, the system can favor whichever action is easiest to generate. Weighting conversion actions by their likelihood of producing value gives the platform a better representation of what the business wants.

Begin with a one-sentence campaign objective:

Acquire the right customer for this offer at an allowable cost, measured by the most reliable purchase, qualified-lead, revenue, or repeat-value signal available.

Then audit every conversion action against that objective:

List every action currently counted in campaign reporting and bidding.
Identify the business outcome that happens after each action: qualification, sale, revenue, retention, or no meaningful progress.
Classify the action as a primary outcome, a useful secondary signal, or a diagnostic event.
Assign relative values only where you can defend the differences with business logic or downstream data.
Remove weak proxy actions from optimization when they compete with stronger outcomes.

Observed action	How to treat it	Question to answer first
Purchase with recorded revenue	Use as a primary value signal when the revenue is reliable	Does revenue reflect the full order without duplicates or missing transactions?
Qualified phone call or sales-ready lead	Weight according to its downstream likelihood of becoming a customer	Can you distinguish a qualified inquiry from support, spam, or a poor-fit prospect?
Unqualified form submission	Keep secondary until qualification data proves its value	What share reaches the next meaningful sales stage?
Page view, content download, or other micro-conversion	Use for diagnosis or audience building, not as a substitute for revenue	Does this action predict a valuable outcome, or is it merely easy to complete?

A phone call isn’t inherently more valuable than a form submission. It deserves more weight only when your own qualification and sales data show that it is more likely to create value. The same rule applies to any conversion hierarchy: evidence should determine the weight, not a generic PPC convention.

Google’s planning direction reinforces the need for clear outcome signals. Performance Planner has stopped supporting Display and Video planning as well as impression-share-based plans, while its supported scope centers on conversion-oriented campaign types such as Search, Shopping, App, Demand Gen, Local, and Performance Max. That doesn’t make awareness activity worthless. It does mean you need your own explanation of what upper-funnel spend contributes instead of treating impressions as sufficient proof.

Don’t invent precise values merely to satisfy an automated system. False precision can redirect real budget. If the downstream value is unknown, preserve the action for reporting, investigate its relationship to sales, and keep the uncertainty visible until you have a defensible signal.

Route each kind of intent to the right campaign treatment

Conversion-focused targeting begins before you select a match type or audience. You need to know what the person is trying to accomplish and how close that intent is to a decision.

For every meaningful query or audience, ask three questions:

Who has a present problem and is likely to act now?
Who could become a buyer after an objection is answered?
Who is unlikely to buy because the offer, use case, price, or customer profile doesn’t fit?

This classification should change the ad, landing page, bidding signal, and degree of structural control. It shouldn’t remain a persona exercise in a planning document.

Use precision where the intent justifies it

High-intent, high-value terms can merit dedicated control. Selective single-keyword ad groups may improve message relevance and query precision where one term represents commercially important demand. That doesn’t justify rebuilding an entire account around single-keyword structures. Reserve the added maintenance for cases in which the intent and potential value make it worthwhile.

Competitor searches can also represent developed purchase intent. The person already understands the category and may be evaluating alternatives. A competitor campaign therefore needs a clear reason to choose your offer and a relevant landing page; a generic page wastes the intent you paid to capture.

Target Impression Share is another deliberate exception. It may support brand defense or visibility on strategically important non-branded terms, but it pursues presence rather than conversion efficiency. Use it only when visibility itself is the stated objective and the business accepts the possible efficiency tradeoff. Don’t present the result as a conventional acquisition win if cost per valuable outcome deteriorates.

Let automation explore inside visible boundaries

Broad match can discover demand you didn’t anticipate, but exploration needs a feedback loop. Combining it with assertive negative-keyword management lets the platform search broadly while you continually shape what qualifies. Several useful PPC tactics, including selective SKAGs, controlled broad match, competitor bidding, conversion weighting, and feed refinement, work because they improve the signals or boundaries around automation rather than rejecting automation outright.

Use this query-review loop:

Inspect the actual search query, not just the keyword that matched it.
Label its intent, customer fit, likely value, and relationship to the offer.
Exclude irrelevant or consistently poor-fit themes with negative keywords.
Move commercially important themes into a more controlled treatment when dedicated ads, bids, or landing pages would change the outcome.
Feed useful language from real queries back into ad copy and landing-page messaging.

Top-of-funnel queries require a different scorecard. They may contribute by building remarketing pools or strengthening audience signals even when their direct conversion rate is weak. Keep that spend identifiable, state the support role in advance, and don’t allow upper-funnel activity to hide inside the economics of high-intent acquisition.

Retargeting audiences can serve as a controlled environment for message and creative tests because those users already have some familiarity with the offer. A winning message can then be tested with colder audiences. Familiarity still changes behavior, so treat the retargeting result as a promising hypothesis rather than proof that the same creative will work everywhere.

Diagnose performance from revenue backward

An analyst traces a connected path backward from a completed purchase through checkout, landing page, search, and an advertising tile.

When performance weakens, broad questions such as why did ROAS fall tend to produce broad answers. Diagnose the chain from the business result backward:

Spend to click to conversion to qualified outcome to sale to revenue to repeat value.

The first broken relationship is usually more actionable than the loudest metric in the interface. Use the following patterns as hypotheses to investigate, not automatic verdicts:

If conversion volume rises while Value/Conv. falls, the account may be finding easier but lower-value customers. Inspect audience, query, product, and order-value mix before celebrating the extra conversions.
If raw leads increase while qualified leads do not, improve the conversion hierarchy and customer filters before buying more traffic.
If qualified lead quality remains stable but sales decline, inspect the landing-to-sales handoff, offer, and downstream process rather than forcing a media-only explanation.
If relevant queries decline, examine match behavior and negatives before rewriting every ad.
If click-through performance improves without a better business result, the new message may be attracting attention without improving buying intent.

This is especially important when B2B and B2C demand overlaps. A campaign may collect many inexpensive consumer conversions while losing the higher-value business buyers it was meant to acquire. In that situation, stronger first-party audience inputs, specific audience segments, and value rules can emphasize B2B intent. That approach has been used to address lagging average order value reflected in Google Ads Value/Conv., but it still requires measurement: targeting a supposedly valuable group doesn’t guarantee valuable orders.

Evaluate economics at the deepest reliable level you possess. For ecommerce, revenue per order is more informative than order count, while contribution after variable costs is more useful than revenue alone when the necessary financial data is available. For lead generation, an expected value model can combine qualification likelihood, close likelihood, and customer economics. Use definitions approved by the people responsible for finance and sales rather than creating a parallel PPC version of profitability.

Customer lifetime value can justify a different acquisition decision from first-order revenue, but only when retention and repeat purchases are observable. Ask why customers stay, what causes another purchase, and which segments actually retain. Don’t raise allowable acquisition costs because an AI tool or a planning assumption produced an attractive lifetime-value story.

When you alter conversion values, audience rules, targeting, or campaign structure, log the change and the intended effect. Avoid simultaneously changing so many decision variables that you can’t tell whether performance moved because of better traffic, a different signal, a new message, or a changed offer.

Use AI to produce testable hypotheses, not synthetic certainty

Generative AI is useful when it helps you ask sharper questions. It can rapidly surface possible emotional triggers, buying-intent segments, objections, lifetime-value ideas, and explanations for weak average order value. Better campaign prompts become more useful as they get closer to a concrete audience, offer, and performance problem.

Use prompts as structured briefs. Supply the offer, intended customer, price context, conversion action, observed performance pattern, and any known constraints. Then ask for hypotheses that can be checked against real query, CRM, sales, or order data.

Purchase intent prompt: Separate the audience into people likely to act now, people who need persuasion, and people who are poor fits. For each group, identify the observable evidence that would confirm or reject the classification.
Emotional context prompt: Identify the fears, frustrations, ambitions, and desired relief that could influence this customer. Distinguish plausible motivations from claims requiring customer evidence.
Objection prompt: Generate three to five credible objections to this offer. For each one, propose a response based on logic, emotion, and proof, but flag any proof the business must substantiate.
Value diagnosis prompt: Given rising conversion volume and falling Value/Conv., propose segment, query, audience, product-mix, and order-value explanations. Rank them by what can be checked with the available data.
Lifetime-value prompt: Explain why a customer might stay, buy again, or expand the relationship. Convert each idea into a retention hypothesis and specify what data would demonstrate that it is real.

The output is not customer evidence. AI can make an unsupported psychological profile sound convincing, invent proof, or favor a neat explanation for a messy performance change. Check proposed motivations against search terms, customer language, objections heard by sales, and observed buying behavior. Delete claims you can’t substantiate.

Turn each surviving idea into a compact experiment card:

Hypothesis: what you believe will change and why.
Audience: the specific intent or customer group being tested.
Variable: the message, creative, landing page, query treatment, audience input, or value signal you will change.
Primary measure: the valuable outcome that determines success.
Guardrails: the quality, cost, average-value, or downstream metrics that must not deteriorate unnoticed.
Decision: what you will scale, revise, or stop after interpreting the result.

A test is useful even when it loses, provided it isolates a meaningful decision. A higher click-through rate with weaker lead quality tells you the message attracted the wrong kind of attention. More conversions with lower order value tells you the platform responded to the signal but the signal didn’t represent enough value. Those are findings you can act on.

Key takeaways

Optimize for the deepest reliable business outcome, not the largest conversion count.
Give different conversion actions different treatment when their downstream value differs.
Apply tight control to commercially important intent and give automated discovery explicit boundaries.
Keep upper-funnel activity visible and judge it by its defined support role, not by impressions alone.
When results weaken, trace the path from revenue backward until you find the first relationship that changed.
Use AI to generate and rank hypotheses, then validate them with customer and performance data.

Start with one campaign, not an account-wide rebuild. Write its economic objective, audit the conversion actions influencing bidding, and inspect which queries or audiences produce the valuable outcome. Make the smallest signal or routing change that addresses the gap, record the expected effect, and let the next decision follow from business results rather than interface activity.

References

April 9, 2026

Modern Marketing Growth Models: How to Choose an Agency

You can hire an agency that improves a channel and still end up with a weaker growth system. Paid media may generate cheaper leads that sales cannot convert. Organic visibility may rise while qualified website visits fall. Marketing may create demand that service and operations are not prepared to support.

The answer is not a longer list of tactics. You need a growth operating model that connects customer states, discovery surfaces, commercial outcomes and decision rights. Once that model is clear, you can judge whether an agency will strengthen it or merely manage part of it.

Replace the single funnel with a growth operating system

Inbound marketing gave teams a coherent sequence: attract an audience, convert visitors and nurture leads. That logic remains useful, but it cannot carry the entire growth plan when discovery, evaluation, conversion and retention happen across different systems.

HubSpot’s shift from INBOUND to UNBOUND reflects growth spanning marketing, sales, service and operations across the customer journey. The important lesson is not the conference name. It is that growth no longer belongs to one function or one acquisition framework.

The old relationship between visibility and traffic is changing as well. An AI-generated answer can satisfy part of a search without sending the user to a website. A prospect can encounter a brand in an AI answer, validate it through search, read customer commentary, click a paid ad later and enter the CRM as direct traffic. A channel report may credit the final interaction while missing most of the journey.

A modern growth model should therefore answer four connected questions:

Model layer	Question to answer	Evidence you need
Commercial outcome	What business result are we trying to change?	A primary outcome, its definition and financial or operational guardrails
Customer state	What must become true for the customer to move forward?	Questions, objections, intent signals and points of friction
Discovery and delivery surfaces	Where can we create, capture, convert or retain demand?	A defined role for search, AI answers, content, paid media, sales and service
Learning loop	How will evidence change the next decision?	An owner, review cadence, decision threshold and change record

If one of these layers is missing, the agency will fill the gap with its own assumptions. A media agency may treat platform revenue as the outcome. An SEO agency may treat rankings as the outcome. A content agency may treat publishing volume as the outcome. Those measures can be useful, but none is a substitute for the business result you hired the partner to influence.

Build the growth brief before you write the agency brief

A team arranges interconnected planning tiles and decision markers during a growth strategy workshop.

An agency request for proposal usually starts with services: SEO, paid search, content, analytics or AI optimization. Start one level higher. Describe the growth constraint first, then determine which capabilities are needed to remove it.

Name one primary outcome. State the business result, not the marketing activity. Pair it with guardrails that prevent a local win from damaging lead quality, margin, retention, brand standards or another important constraint.
Map the customer states. Identify what customers need when they are recognizing a problem, evaluating options, making a purchase, adopting the product and deciding whether to continue. Use the states that fit your business instead of forcing every journey into a generic funnel.
Locate the actual constraint. Determine whether the problem is insufficient demand, poor discovery, weak consideration, conversion friction, slow sales follow-up, onboarding failure or low retention. Do not commission more acquisition work when the binding constraint sits after acquisition.
Assign a job to every surface. Decide whether each channel is meant to create demand, capture existing demand, answer a question, support evaluation, convert intent or retain a customer. A surface can support several jobs, but it should have one primary role in the plan.
Define the learning loop. Record what will be observed, who interprets it, which decision it informs and who can approve the change. Reporting without a decision path produces dashboards, not growth.

This is especially important for SEO, answer engine optimization and generative engine optimization. They overlap, but they are not interchangeable line items. SEO can improve discoverability in conventional search. AEO can make an answer easier to extract and present. GEO can focus the work on how generative systems understand, retrieve and represent a brand. Your measurement plan should preserve those distinctions while connecting them to the same customer journey.

Do not force every visibility signal into an immediate revenue calculation. A metric can guide optimization without proving causal impact. Rankings, answer inclusion, brand mentions and qualified visits can show whether discovery is changing. CRM progression, revenue and retention can show whether commercial performance is changing. The agency should explain the relationship between those layers without pretending that one attribution model observes the entire journey.

Your completed growth brief can be one page. It should contain the primary outcome, guardrails, constrained customer state, surface roles, measurement definitions and unresolved questions. That page gives every prospective agency the same problem to solve and makes proposals easier to compare.

Divide ownership before you evaluate capabilities

A growth partner needs room to make decisions, but outsourcing execution does not transfer accountability for the business. Clarify what the brand owns, what the agency owns and what must be shared before discussing deliverables.

The brand should retain business truth. This includes commercial priorities, customer definitions, approved claims, margin constraints, risk tolerance and the final authority over budgets and data access.
The agency should own recommendations and agreed execution. It should identify opportunities, explain trade-offs, perform work within the approved boundaries and maintain a record of material changes.
Measurement should be shared. The agency may build reports, but metric definitions, attribution limitations and tracking changes must be visible to both sides. Neither party should be able to change the meaning of success silently.
Cross-functional decisions need one accountable lead. Someone must reconcile conflicts among marketing, sales, service and operations. A committee can contribute, but it cannot substitute for a named decision-maker.

This ownership map also exposes misleading claims of being full service. A long service menu tells you what an agency is willing to sell, not where it repeatedly performs strong work. Ask what percentage of clients actually use each advertised service. Then ask who leads that work, what other capability it depends on and where the agency normally brings in outside expertise.

Build a simple capability map for every service that matters to your brief. Record the service, client utilization, named practice lead, proposed account owner, proof artifact, dependencies and known limitations. A strong specialist can be a better fit than a nominally full-service agency if your team is prepared to integrate the work. A broad partner can be the better choice when coordination is the main constraint. The right answer depends on the operating model, not the size of the service catalog.

Audit the agency’s decisions, not its pitch language

Client and agency leaders evaluate branching decisions and trade-offs while an abstract presentation remains in the background.

Most agencies can produce a polished audit and a plausible list of opportunities. Your evaluation should reveal how the team prioritizes, measures, automates and changes course after the pitch is over.

Ask six questions that require operational answers

Which services are genuinely central to your business, and what percentage of clients use each one? Look for a precise denominator, a distinction between core and occasional work, and a candid explanation of where the agency is not the best fit. A service list with no utilization data does not establish depth.
How do you combine platform automation, AI optimization and human judgment? Ask which decisions are delegated to platforms, which inputs the team controls, which guardrails prevent undesirable optimization and what triggers human intervention. “AI-powered” is a label, not an operating procedure.
How does reporting lead to a decision? Have the team walk through an anonymized reporting environment. Ask them to start with the business outcome, trace the supporting indicators, identify an uncertainty and show the action that followed. Revenue and return on ad spend may belong in the view, but the team should also explain attribution assumptions and data limitations.
Who will work on the account, and what is the team’s relevant industry tenure? Get names, roles, responsibilities and escalation paths. Distinguish the senior experts who appear in the pitch from the people who will perform and review the work.
How does your team use generative AI on client work? Separate internal uses, such as analysis or drafting, from advertising-platform automation. Ask which client data can enter a tool, what receives human review, how outputs are checked and how material decisions are documented.
What would you inspect first to reduce waste without suppressing growth? A strong answer should describe a sequence: validate measurement, preserve a baseline, inspect settings and allocation, identify suspected waste, estimate the downside of a change and verify the effect after implementation. A promise to cut spend immediately is not evidence of efficiency.

Score each answer from zero to two. Give zero for a vague claim, one for a credible process without supporting proof, and two for a specific process backed by an artifact and a named owner. This produces a maximum score of 12, but the total is less important than the pattern. A partner that scores well on capabilities but poorly on measurement or ownership can create activity faster than it creates learning.

Set knockout conditions before the presentations begin. Examples include refusing to identify the delivery team, being unable to explain data handling, treating platform-reported attribution as unquestionable, or requesting unrestricted budget authority before measurement is validated. Predefined conditions prevent presentation quality from overriding operational risk.

Turn the winning answers into the working agreement

Anything important enough to influence agency selection belongs in the operating agreement. Otherwise, the senior strategist, reporting method or review practice that won the pitch may disappear during delivery.

Decision rights: Record who can change budgets, targeting, conversion events, content claims, schema, site templates and measurement configurations.
AI boundaries: Define approved uses, prohibited data, review requirements and the person accountable for an AI-assisted output.
Change control: Preserve the baseline, document material changes and record the expected effect before implementation.
Reporting logic: Require each review to show what changed, how confident the team is, what may have caused it, what decision follows and who owns that action.
Escalation: Specify what happens when tracking fails, automation pursues the wrong signal, spend moves outside an agreed boundary or results conflict across systems.
Capability continuity: Define how staffing changes are communicated and how critical account knowledge is transferred.

Give a new partner read access before authorizing material changes whenever the platform permits it. Validate conversion definitions, tracking and historical baselines first. Changing optimization events and budgets at the same time can make the result difficult to interpret, and automation can scale the wrong objective quickly. The safer sequence is to establish measurement, document the hypothesis, make a bounded change and inspect the result before expanding it.

The same discipline should continue after onboarding. Do not evaluate the relationship by deliverable volume alone. Evaluate whether the agency is improving decision quality: finding the real constraint, making uncertainty visible, reducing waste, connecting work across the journey and leaving your team with a clearer understanding of what to do next.

Key takeaways

A modern growth model connects commercial outcomes, customer states, discovery surfaces and a defined learning loop.
Write the growth problem before selecting services. Otherwise, every agency will frame the problem around what it sells.
Keep business truth and final accountability with the brand while giving the agency explicit execution and recommendation rights.
Test full-service claims with client utilization, named specialists, dependencies and proof of repeatable delivery.
Evaluate platform automation and internal generative AI separately; both require clear inputs, guardrails, review and escalation.
Convert important pitch promises into decision rights, reporting rules, staffing commitments and change-control procedures.

Before your next agency conversation, complete the four-layer growth model for one important constraint and send the six audit questions in advance. Ask every contender to answer with artifacts, named owners and explicit limitations. The partner that can work inside that level of clarity is far more useful than one that merely offers the longest list of channels.

References

April 8, 2026

AI-Driven Paid Acquisition: A Lead Generation Playbook
If AI-led campaigns keep producing form fills that sales rejects, the system may be succeeding at the wrong task. A thank-you page tells an ad platform that an action occurred. It does not tell the platform whether the lead was qualified, reachable, commercially relevant, or likely to become revenue.

Your first job is to connect those business outcomes to acquisition. Your second is to make the offer equally clear on the landing page, in the feed, across map profiles, and inside every creative asset. Do those two things before increasing spend, and automation has a much better signal to optimize.

Key takeaways: what to fix before spending more
- Optimize toward business quality, not raw form volume. Define an accepted lead, return downstream statuses from the CRM, and keep diagnostic actions separate from primary conversion goals.
- Make the offer unambiguous. A visitor and an automated system should both be able to identify what you sell, who it is for, why it matters, what action to take, and what happens next.
- Measure each funnel stage on its own terms. Awareness, consideration, lead capture, qualification, opportunity creation, and revenue do not share one useful success metric.
- Treat feeds, map listings, structured data, pages, and creative as one information system. Conflicting names, categories, locations, or conversion labels weaken both targeting and attribution.
- Audit placements as well as campaigns. Automated campaigns can reach visual discovery surfaces that behave differently from conventional text search, so a blended click-through rate can hide what changed.
Teach the buying system what a qualified lead means

Begin in the CRM or lead management system, not in the bidding interface. Write down the point at which an inquiry becomes worth pursuing. That definition might depend on service fit, geography, budget, need, or another criterion your sales team already uses. The exact criteria are yours; the important part is that marketing, sales, the CRM, and the ad platform use the same definition.

Then trace the feedback loop:
<!– wp:list {
April 7, 2026

How to Control Automated Paid Search for Commerce Growth

You did not lose control of paid search when platforms automated bidding, audience expansion, and ad assembly. Control moved upstream. The expensive mistake is still managing the account as though a perfect keyword list can compensate for weak conversion data, muddled economics, thin creative, or a poor product page.

Your job now is to give the system a clear commercial objective, reliable evidence, and firm boundaries. Do that well and automation can explore more demand than a person could manage manually. Do it poorly and it will scale the wrong outcome with impressive efficiency.

Control the system through the inputs it learns from

Keywords still matter, but they no longer carry the account on their own. In automated search, keywords function alongside conversion data, first-party audience information, creative assets, and landing-page content. The practical shift is simple: your campaign structure is no longer the whole strategy. It is one part of the training environment you create for the platform.

That is why an automation feature should never be evaluated only by whether it finds additional conversions. Some AI Max campaigns have been credited with up to 27% more conversions, but that is a reason to run a controlled test, not a forecast you should put into a budget. More conversions help only when they are valid, incremental enough to matter, and economically acceptable.

Control area	Decision you own	Evidence to inspect
Business outcome	Which conversion is primary and how it is valued	Completed orders, revenue, margin proxy, cancellations, and returns
Learning data	Which customer and transaction signals are accurate enough to use	Duplicate events, missing values, currency consistency, and match quality
Demand	How discovery traffic is separated from proven demand	Search terms, product-level sales, conversion rate, ROAS, and ACOS
Experience	Which product information, creative, and destination represent the offer	Message continuity, availability, price, page relevance, and purchase completion
Risk	Where automation may spend and when a person must intervene	Budgets, exclusions, brand traffic, inventory, and unexplained mix changes

Start with a conversion contract: a short, explicit definition of what the bidding system is supposed to maximize. This is not a tracking implementation document. It is the agreement between marketing, commerce, and analytics about what counts as success.

Name the primary event. For a commerce campaign, that will usually be a completed purchase. Add-to-cart, product-view, and checkout events can remain useful diagnostics without being treated as equivalent to revenue.
Define the value. Decide whether the platform receives gross order revenue, a margin-weighted value, or another consistent commercial proxy. If two orders produce very different contribution margins, equal revenue values may teach the system to prefer the less profitable mix.
Define validity. Document how duplicate purchases, cancellations, refunds, taxes, shipping, and currency are handled. A bidding model cannot infer that an inflated or duplicated value is wrong.
Define the observation window. Review performance only after the normal conversion and reporting lag has had time to mature. Otherwise, recent traffic will look artificially weak and invite unnecessary changes.
Name an owner. Someone must be accountable for detecting broken events, abrupt value changes, and gaps between platform reporting and the commerce system.

Well-structured first-party data now does much of the strategic work once associated with exhaustive keyword research. It helps the platform distinguish valuable customers and transactions from activity that merely looks busy. But volume does not cure bad measurement. A larger stream of duplicated purchases is still bad data, and automation can magnify its effect faster than a manual bidder would.

Before expanding automation across the account, validate the contract in a bounded campaign or product group. Changing conversion definitions, bidding targets, audience inputs, and creative at the same time can expose the business to avoidable spend while making the result impossible to interpret.

Separate discovery from profitable scale

An exploration area tests many generic products while a gated passage leads selected products into orderly fulfillment lanes.

Commerce advertising has two jobs that pull in different directions. Discovery needs freedom to test unfamiliar queries, audiences, and products. Performance needs concentration: more budget behind combinations already linked to acceptable sales. Put both jobs in one undifferentiated campaign and the blended result hides what each dollar is doing.

A stronger architecture creates a deliberate path from exploration to scale. Search environments are especially useful here because shoppers express intent in their queries, while Google Shopping and Amazon Ads can connect that demand to product-level or keyword-level revenue. That creates a feedback loop between search behavior, sales, and budget allocation.

Discovery captures uncertainty. It explores a wider set of eligible demand under its own budget and economic limits. Its purpose is to find useful search terms and product-demand combinations, not to look as efficient as a mature campaign.
Performance concentrates evidence. It gives proven converters dedicated budgets and targets so they do not have to compete with every exploratory term for spend.
Brand protection isolates known demand. Branded searches often behave differently from generic acquisition. Separate reporting prevents strong brand results from disguising weak prospecting.
Ranking activity has an explicit cost. If you spend more aggressively to improve visibility or marketplace position, keep that objective distinct from a profit-maximizing campaign.

The handoff between discovery and performance should use written promotion rules. A term or product is not proven because it converted once, and it should not stay in discovery forever after building credible evidence. Define the minimum evidence your business needs, then test that evidence against four questions:

Has the query or product produced enough mature sales to reduce the chance that one unusual order controls the decision?
Does its ROAS or ACOS fit the contribution economics of that product after the costs the business actually bears?
Can inventory and fulfillment support more demand without creating cancellations or a poor customer experience?
Does the landing page or marketplace listing genuinely satisfy the intent that generated the sale?

Use demotion rules as well. A proven term can return to discovery or lose budget when its economics deteriorate after a mature measurement window, when stock becomes unreliable, or when the offer no longer matches the query. Graduation is a status based on current evidence, not a permanent award.

Do not impose one universal efficiency target on every layer. Discovery may operate under a stricter spending cap while accepting more variance. A performance campaign may receive more budget but face a firm profitability requirement. Brand and ranking campaigns need their own definitions of success. The crucial point is that each layer has a known job, budget, and exit condition.

Use platform-specific structures without losing the common logic

Google Shopping and Amazon Ads can share the same discovery-to-scale strategy, but their campaign mechanics and commercial roles are different. Reproducing the same campaign map on both platforms creates superficial consistency at the cost of useful control.

Route Google Shopping demand through distinct layers

A workable Google Shopping structure uses three layers: a branded layer, a catch-all discovery layer, and a dedicated layer for the strongest terms. Campaign priority and other routing controls can then help prevent exploratory demand from consuming the budget reserved for proven opportunities.

Branded layer: A shopping-focused, assetless Performance Max campaign can be used to concentrate on shopping inventory and reduce unintended expansion into other channels. Inspect the actual traffic and placement mix rather than assuming the setup label guarantees isolation.
Catch-all layer: Keep a wide net for search-term discovery, but contain it with a separate budget and lower bids or a suitably conservative target. Its output is evidence: which queries and products deserve focused investment.
Performance layer: Move reliable, high-intent demand into a dedicated campaign where budget and bidding can reflect its demonstrated economics.

This structure is useful only if routing works as intended. Inspect search terms, product distribution, brand share, and channel mix. If the catch-all keeps taking proven demand, or the branded layer expands beyond its assignment, the labels on the campaigns are not describing the account you actually have.

Performance Max can also operate alongside AI Max for Search, but overlap should have a reason. Decide which campaign is responsible for known product demand, which is exploring broader intent, and how you will detect duplication or channel substitution. Reach is not automatically incremental growth.

Organize Amazon Ads around the SKU and the commercial objective

Amazon gives you a different feedback loop. The shopper is already in a marketplace, reporting can be granular at the product and category level, and ad conversion can contribute to stronger organic position. The practical structure is therefore SKU-level research, performance, and ranking tiers.

Research tier: Explore broad keyword possibilities and collect evidence about how shoppers describe the need. Control the downside with a defined budget and ACOS boundary.
Performance tier: Concentrate proven converters and manage them toward the product’s profit requirement.
Ranking tier: Bid more aggressively only when improving organic position is a deliberate objective and the business has approved the cost of doing so.

ROAS and ACOS describe the same relationship from opposite directions. ROAS is attributed revenue divided by ad spend. ACOS is ad spend divided by attributed revenue. Neither metric knows your profit. Set the acceptable range from contribution margin after relevant product costs, marketplace fees, fulfillment, discounts, and expected returns. A generic benchmark can make an unprofitable SKU look healthy or constrain a high-margin SKU that could support more growth.

Higher conversion rates on Amazon can support organic ranking and reduce later acquisition pressure, but do not count that future benefit twice. Keep direct ad economics visible, document when ranking is the primary objective, and check whether organic position actually changes before continuing the extra spend.

Across Google and Amazon, use the same product economics as the common language. The campaigns may optimize differently, but both should ultimately answer whether the next unit of spend creates acceptable commercial value.

Make product data, creative, and landing pages part of targeting

When automation assembles ads and expands matching, every customer-facing input can affect both eligibility and persuasion. Creative is not decoration added after targeting. Landing-page content is not merely the place traffic goes. These assets help the system interpret what you sell, who may want it, and which message belongs with a particular intent.

Build a message system for each important product group before asking the platform to generate combinations. It should cover:

Product identity: What the item is, using the language a qualified shopper would recognize.
Use case: The job, occasion, or problem the product genuinely addresses.
Differentiator: A factual reason to choose it over a plausible alternative.
Proof: Verifiable product details, policies, or other substantiation available on the destination.
Offer conditions: Price, eligibility, availability, shipping, or promotional limits that could change the buying decision.

That framework gives automation useful variety without inviting random claims. It also makes creative testing interpretable. If one asset emphasizes a use case and another emphasizes price, you can learn something from the difference. If every asset changes the product, audience, offer, and tone at once, a winning combination tells you little about why it worked.

Then audit continuity from query to ad to destination. A shopper who searches for a specific variant should not land on a generic category page and be expected to restart the search. A promotion in an ad should be visible with the same conditions on the page. Product names, images, price, availability, and purchase options should agree across the feed, creative, and destination.

Landing-page quality matters twice. It affects whether a visitor can complete the purchase, and automated systems can use the post-click experience and page content as relevance signals. Diagnose a weak product group accordingly. The problem may be bidding, but it may also be a page that sends an ambiguous signal or fails to finish the promise made by the ad.

Confirm that the destination resolves to the correct product or tightly matched category.
Keep price, inventory, variant, and promotion information synchronized with the advertisement.
Make the primary purchase action obvious and functional on the devices receiving paid traffic.
Remove claims from generated or assembled creative when the destination cannot substantiate them.
Separate products with materially different margins, availability, or buying intent instead of forcing them into one undifferentiated asset and bidding group.

Do not compensate for a weak offer with broader automation. Broader matching can find more people, but it cannot make an unclear product, unavailable variant, or contradictory price more attractive. Fix the commercial experience before paying the system to expose it at greater scale.

Run a human operating system around the automation

Four professionals surround a circular control table, reviewing product, creative, storefront, and conversion inputs around an automated sorting mechanism.

The human role is not to outbid the bidding model one adjustment at a time. It is to decide what the model should learn, recognize when the evidence has become unreliable, and intervene at the level that caused the problem.

Use a repeatable review loop:

Observe mature performance. Wait for the normal reporting and conversion lag, then compare actual results with the campaign’s stated job.
Locate the failure class. Check measurement, demand mix, product economics, inventory, creative, destination, and campaign routing before changing bids.
Change one class of input. For example, repair conversion values, adjust a budget boundary, refine routing, or replace weak assets. Avoid simultaneous changes that erase causal clarity.
Write the expected effect. Record what should change, which metric should reveal it, what observation window is appropriate, and what would justify reversal.
Promote, hold, demote, or stop. Use the rules established for discovery and performance rather than making a fresh subjective decision every time.

Not every bad-looking period calls for intervention. Hold when conversion data is still immature and spend remains inside the approved boundary. Change the campaign when mature evidence shows a persistent problem with an identifiable input. Stop or contain it immediately when tracking breaks, spend escapes its guardrail, inventory cannot support orders, or an ad makes an inaccurate claim. Those failures can waste money or harm customers while the model continues optimizing against corrupted conditions.

Your review should also distinguish a performance change from a mix change. A stable blended ROAS can conceal a shift from new-customer demand toward branded traffic, from high-margin products toward low-margin products, or from direct shopping placements toward less valuable inventory. Look below the account total before calling automation successful.

Keep an intervention log. For every material change, record the campaign, business reason, affected products, input changed, expected outcome, and rollback condition. This turns account management into an accumulating decision system instead of a sequence of reactions. It also prevents one operator from undoing another operator’s test without knowing why it exists.

Key takeaways

Keywords remain useful signals and diagnostics, but conversion quality, first-party data, creative, and landing pages increasingly determine what automated campaigns learn.
Define the primary conversion, its value, its validity rules, and its owner before expanding automation.
Give discovery, proven performance, branded demand, and ranking activity separate jobs, budgets, and exit conditions.
Use the same discovery-to-scale logic across Google Shopping and Amazon Ads, but adapt the campaign mechanics to each platform.
Judge ROAS and ACOS against product contribution economics rather than a generic account benchmark.
Let people own measurement, commercial judgment, guardrails, creative truth, and the decision to promote or stop an experiment.

Start with one meaningful product group. Write its conversion contract, calculate its acceptable economics, identify which traffic is discovery and which is proven, and audit the message from query through purchase. Only then widen automation. If you cannot explain the value entering the bidding system, the system is not ready to scale it.

References

April 3, 2026

How to Make LinkedIn Recruitment Campaigns More Efficient

Your LinkedIn recruitment campaign can generate plenty of clicks and applications while still failing at the one outcome that matters: producing qualified hires at a sustainable cost. When interview volume stays flat as campaign activity rises, you are probably paying for attention rather than candidate fit.

The remedy is not simply a narrower audience or a lower bid. You need a campaign system that identifies intent, filters candidates before expensive actions, separates different stages of demand, and connects media spend to interviews and hires.

Define efficiency before you buy another click

Recruitment efficiency is not a high click-through rate, a cheap click, or even a low cost per application. Those metrics describe parts of the journey. They do not tell you whether the campaign is helping the company hire suitable people.

Start with a complete conversion chain. Every active campaign should be traceable through these stages:

Ad click or lead interaction.
Pre-qualification page visit.
Application start.
Completed application.
Qualified application.
Interview.
Hire.

Define a qualified application with the hiring team before launch. It might require a particular certification, a minimum level of relevant experience, permission to work in the required location, or another genuine condition of the role. If recruiters apply different definitions after applications arrive, campaign comparisons will be unreliable.

Calculate cost per hire using one consistent scope: the spend assigned to a campaign divided by the hires attributed to it. If you include creative, agency, or platform costs, include them consistently across every campaign you compare. Apply the same attribution rule as well. A neat dashboard cannot rescue inconsistent definitions.

Your working report should show spend, clicks, completed applications, qualified applications, interviews, and hires for each campaign. Add conversion rates and costs between stages. That makes the source of waste visible:

High click-through rate but few applications: the ad may be creating curiosity that the role cannot satisfy, or the application handoff may be too demanding.
Many applications but few interviews: your audience, creative, or landing page is not doing enough pre-qualification.
Qualified applicants and interviews but few hires: inspect the offer, recruiter follow-up, interview process, and hiring decision before changing the ads.
Hires from one segment but weak volume: increase that segment carefully instead of loosening the requirements across the whole account.

The first two patterns are especially important because click and application volume can conceal poor alignment. Optimizing to the earliest available event encourages the campaign to find more of that event, not necessarily more people the hiring team wants to meet.

For early testing, manual cost-per-click bidding can give you tighter control over how quickly the budget is exposed. Consider automated bidding after conversion tracking is working and the campaign has produced a stable enough mix of qualified applicants to judge. The purpose is not to defend manual bidding forever. It is to avoid paying an automated system to amplify an unproven audience or message.

Build audiences from fit and intent, then keep them separate

Diverse professionals move along separate teal and amber pathways while a translucent lens highlights people where fit and intent overlap.

Job title, industry, and seniority tell you who a person is professionally. They do not tell you why that person might consider changing jobs. A more useful audience plan combines three layers:

Core fit: relevant titles, skills, certifications, and experience.
Behavioral intent: open-to-work status, recent job-seeking activity, relevant group membership, or engagement with industry content, where those signals are available in your campaign setup.
Career-friction hypotheses: roles associated with burnout, employers affected by layoffs, or environments where advancement may be limited.

Use career friction to form a messaging hypothesis, not to pretend you know how an individual feels. An employee at a competitor is not automatically dissatisfied. A person in a demanding profession is not automatically burned out. Your ad can describe a credible alternative without making a personal claim about the viewer.

Give each intent level its own campaign job

Active candidates and cold passive candidates should not share the same budget, message, and success expectation. Separate them so that a high-intent audience cannot hide waste in a broad awareness campaign.

Intent segment	Useful audience signals	Message	Campaign job
High intent	Open-to-work users, recent job seekers, and retargeting audiences	Role specifics and a direct application invitation	Generate qualified applications now
Warm passive	Relevant skills, competitor employers, and niche professional groups	A concrete career, schedule, compensation, or lifestyle improvement	Turn openness into consideration
Cold passive	Broader qualified audiences and lookalike audiences	Employer reputation, culture, mission, and realistic day-in-the-life content	Build a future talent pool

This high-, warm-, and cold-intent structure also changes how you interpret performance. A cold employer-brand campaign should not be expected to match the immediate application rate of retargeting. Its job is to create an audience that a later campaign can convert more economically.

Control overlap when you build these segments. Start with the most specific high-intent pool, then exclude it from warm campaigns where your setup allows. Exclude both from the cold campaign. Without those exclusions, the same promising candidate can appear in several campaigns, making cost and conversion comparisons harder to trust.

Skill-based segmentation is often more actionable than one large professional audience. If a role accepts candidates from several disciplines, place each major skill group in a separate campaign and adapt the value proposition. You will see which background produces qualified applicants, rather than averaging unlike candidates into one result.

Make the ad qualify candidates before they click

A recruitment ad has two jobs: attract the right person and discourage the wrong person from spending your budget. If the ad hides hard requirements to maximize clicks, the application process has to reject those people later, after you have paid for their attention and consumed recruiter time.

A practical recruitment ad contains four elements:

A recognizable identity or friction: name the professional situation the role improves.
A hard fit statement: specify the required role, skill, certification, or experience.
A verified reason to move: state the real compensation, flexibility, schedule, growth path, mission, or working conditions.
A clear boundary: say when the position is not entry-level or requires a specific background.

Use this fill-in structure when drafting creative:

[Professional identity]: If [specific, credible friction] is making you consider a change, [company] is hiring for [role]. You will need [must-have requirements]. The position offers [approved and verifiable benefits]. This role is not suitable for [clear exclusion]. [Direct next step].

The exclusion is not an apologetic footnote. It is part of the offer. Phrases such as “requires enterprise account management experience” or “not an entry-level position” can reduce irrelevant responses and protect recruiter capacity. The same principle applies to licensed or specialist roles: put the non-negotiable credential in the ad, not halfway through the application.

Only promote benefits the employer has confirmed. “Flexible schedule” is not useful filtering language if flexibility depends on the manager. A compensation claim should match the actual structure and conditions. An exaggerated promise may raise clicks, but the mismatch will surface in application abandonment, interviews, or offer rejection.

Test the message against qualified outcomes

Run creative tests that change one decision-relevant element at a time. You can compare an identity-led opening with a friction-led opening, test schedule against career growth as the primary value proposition, or move the hard qualification earlier in the copy. Keep the audience, role, and destination consistent while you test.

Do not declare a winner because one variation earns more clicks. Compare completed applications, qualified-application rate, interview rate, and eventual hires. The more selective ad may have a lower click-through rate and still be the more efficient recruitment asset.

For specialized or senior positions, a narrowly targeted Message Ad can carry more context than a short feed ad. Keep the outreach specific and easy to decline:

Hi [First Name], your background in [relevant skill or field] stood out. We are hiring a [role] for people with [must-have experience]. The position offers [two verified benefits], and it is intended for [seniority or specialist profile], not entry-level candidates. Would you be open to a brief conversation? If not, thank you for considering it.

Broad message campaigns can become expensive quickly. Reserve this format for audiences whose eligibility and likely value proposition are already well defined.

Use a two-stage application path and retarget real interest

A job seeker begins on a smartphone, passes through a qualification gateway, and reaches an interview table while glowing connections loop back to other interested candidates.

Sending every click directly to a long applicant-tracking form forces candidates to do too much before they understand the role. It also prevents you from distinguishing between a poor offer and a difficult application experience.

Use a two-stage path instead:

Pre-qualification page: explain the work, expectations, location or schedule, compensation details, must-have criteria, and who should not apply.
Short application: ask only for the information needed to evaluate the next step, or use LinkedIn Easy Apply when it suits the hiring workflow.

The first stage should increase clarity, not create an obstacle course. A reported 30-50% reduction in cost per hire has been associated with this two-step structure, but treat that range as a directional campaign claim rather than a forecast. Your result will depend on the role, offer, audience, tracking, and existing application process.

Instrument both stages separately. Track the proportion of ad visitors who reach the page, start the application, complete it, qualify, interview, and get hired. If many suitable-looking visitors leave before starting, inspect the offer and page. If many begin but do not finish, inspect the form. If completions are high but interview selection is low, strengthen the qualification language.

Retarget people according to what they already did

Not every qualified person applies during the first visit. Build retargeting audiences from career-page visitors, ad viewers, and people who watched at least 50% of a recruitment video. Their next message should move the decision forward rather than repeat the original ad.

Career-page visitor: restate the role’s main benefit and the most important qualification.
Substantial video viewer: show an employee outcome, realistic role detail, or day-in-the-life proof that answers a likely concern.
Application visitor who did not complete: return to the role and a shorter next step, if your tracking and campaign rules support that audience.
Interested candidate near a genuine deadline: communicate the real closing date. Do not manufacture urgency.

Exclude people who have already applied unless the follow-up has a deliberate recruiting purpose. Otherwise, you keep paying to ask for an action they have completed and distort the apparent efficiency of the retargeting campaign.

Once the core funnel is working, expand carefully. Competitor-employee targeting can emphasize a verified advantage without attacking another employer. Skill-specific campaigns can reveal which backgrounds convert. Targeted messages can reach a small pool of senior specialists. Each tactic should remain separate enough that you can identify its qualified applications, interviews, and hires.

Key takeaways for your next recruitment campaign

Measure cost per qualified application, interview, and hire alongside clicks and completed applications.
Define qualification with recruiters before launch so campaign comparisons use the same standard.
Combine core professional fit with available intent signals instead of targeting job titles alone.
Separate high-intent, warm passive, and cold passive candidates because they need different messages and success criteria.
Put must-have requirements and meaningful exclusions in the ad to prevent avoidable clicks.
Use a clear pre-qualification page followed by a short application, then track the handoff between them.
Retarget demonstrated interest with a next-step message and exclude candidates who have already applied.
Move budget according to qualified applications, interviews, and hires, not the campaign with the busiest top-line metrics.

Before increasing your next LinkedIn budget, rebuild one role from end to end. Separate active and passive audiences, add one hard qualifier to the creative, route candidates through a concise role page, and add qualified applications, interviews, and hires to the campaign report. That smaller redesign will show you where the waste actually begins.

References

CrushPress.AI – Maximize Hiring Efficiency: Lower Costs with LinkedIn Campaigns

April 1, 2026

Performance Max Campaign Controls: A Practical Playbook
You do not need complete control of Performance Max to keep it accountable. You need to know which reports merely describe what happened, which settings impose hard limits, and which inputs steer the automation without guaranteeing an outcome.

The most reliable approach is to work in that order: verify what the campaign is optimizing for, remove clearly unwanted traffic, apply narrow constraints where the evidence is strong, and then improve the creative, feed, budget, and bidding inputs. That gives you more control without excluding useful demand just because a report looks uncomfortable.

Key takeaways
- Campaign-level negative keywords, placement exclusions, ad schedules, demographic exclusions, and device controls are the clearest direct controls available in Performance Max.
- A report is not automatically a control. Search terms can lead directly to negatives, but placement impressions do not tell you how much a placement spent or whether it produced conversions.
- Use exclusions for traffic that is demonstrably irrelevant, ineligible, unsafe for the brand, or operationally impossible to serve. Do not use them as a reflex whenever performance is uncertain.
- Creative assets, product feeds, conversion goals, bids, and budgets steer where automation looks for results. They usually deserve attention before you start narrowing reach aggressively.
- Record each material change and its reason. If you change negatives, schedules, devices, assets, and bidding together, the next report cannot tell you which decision helped.
Remove obvious waste with search terms and placement controls

The safest exclusions begin with a simple question: could this traffic ever produce the outcome you want? If the answer is clearly no, blocking it protects the budget. If the answer is merely uncertain, investigate before turning an observation into a permanent rule.

Turn search-term visibility into a disciplined negative list

Campaign-level negative keywords can be added from the Performance Max search terms report. This removes much of the friction that once separated finding an irrelevant query from blocking it.

That convenience makes restraint more important. A query with no recorded conversion is not automatically irrelevant. It may have appeared too infrequently to judge, sit earlier in the buying journey, or suffer from a landing-page or offer problem. Negatives should remove unwanted meaning, not conceal a broader performance issue.

Use this review sequence:
1. Group terms by intent rather than reacting to isolated wording. Repeated patterns reveal more than one unusual query.
2. Separate clearly impossible or irrelevant intent from ambiguous intent. Exclude the first group; investigate the second.
3. Check whether a candidate negative could also match valuable searches. Use the narrowest exclusion that removes the unwanted concept without cutting into legitimate demand.
4. Add the negative from the search terms report and record why it was added. A short reason makes later reversals much easier.
5. Review the effect in the next stable comparison period, allowing for the conversion lag that normally applies to your account.
Common candidates include searches for a service you do not provide, a product category you do not sell, or an intent that cannot become a qualified customer. A merely expensive term belongs in a different bucket. Before excluding it, check the conversion goal, landing page, offer, and query context.

Use placement data for suitability before profitability

Performance Max placement visibility now sits in the campaign’s expanded reporting and exclusion workflow, including the ‘Where ads have shown’ area. The placement report is particularly useful for spotting large volumes of impressions in contexts that do not fit the campaign, such as unintended mobile apps or children’s programming.

The limitation matters: impression-level placement data is not a placement-level profit-and-loss statement. A placement with many impressions has not necessarily consumed an equivalent share of spend, generated the same share of clicks, or caused the campaign’s overall inefficiency. Treating impressions as cost can lead you to exclude inventory for the wrong reason.

Placement exclusions are strongest when the decision is about relevance or brand suitability. If a context is plainly inappropriate, an account-level negative placement may be justified. Because that scope can affect more than the campaign you are reviewing, check which other campaigns rely on the same inventory before applying it.

If the concern is performance rather than suitability, look for corroborating evidence first. Review the campaign’s search intent, channel distribution, assets, conversion goals, and landing pages. The placement report may identify where to investigate, but it does not always identify what to remove.

Apply time, demographic, and device limits without choking reach

Schedules, demographic exclusions, and device settings are genuine constraints. They can improve efficiency when they reflect how the business actually operates. They can also starve the campaign when they are used to compensate for weak data, a broken experience, or impatience with normal variation.

Build an ad schedule around opportunity and operating capacity

The ‘When and where ads showed’ reporting area provides hour-by-hour information even when the campaign began without a restricted schedule. You can apply a schedule under ‘Campaigns > Audiences, keywords, and content > Ad schedule’.

Scheduling is most useful when budget is limited and there is a repeatable mismatch between ad delivery and the business’s ability to convert demand. A lead-driven company may struggle to handle inquiries during certain hours. A campaign with a constrained daily budget may spend during weak periods and lose access to stronger periods later. In either case, the schedule should reflect a demonstrated operating constraint, not a single quiet hour in a report.

Before removing an hour or day, ask three questions:
- Does the pattern repeat across comparable periods, or is it driven by one unusual day?
- Was there enough activity to make the absence of conversions meaningful?
- Could conversion lag, offline follow-up, or the sales process make the hour look weaker than it really is?
If those checks support the same conclusion, restrict the weakest period first rather than rebuilding the entire week at once. A narrow change preserves more eligible inventory and gives you a cleaner result to evaluate.

Reserve demographic exclusions for durable mismatches

Campaign-level demographic exclusions are available under ‘Other settings’. They are appropriate when a group cannot reasonably use or qualify for the offering, or when a consistent body of campaign evidence supports the restriction.

A weak short-term result is not the same as a durable mismatch. Demographic segments may receive different volumes and enter at different points in the customer journey. If you exclude a segment after a small amount of activity, the campaign loses the chance to learn whether better creative, a different landing page, or more complete conversion data would change the result.

Use demographic controls as eligibility rules first and optimization rules second. When the decision is performance-based, document the evidence and plan a later review. An exclusion should remain reversible when the underlying audience or offer could change.

Diagnose the device experience before excluding the device

Device controls in ‘Other settings’ let you review which devices contribute to campaign goals and decide which devices to include or exclude. This is valuable, but device performance often exposes a site or journey problem rather than an audience problem.

Before excluding a device, complete the conversion path on that device. Check whether the page loads cleanly, forms are usable, calls work, product information remains legible, and the final action can be completed without friction. If the experience is broken, repair it. Excluding the device may reduce visible waste, but it also hides the defect and abandons otherwise valid demand.

A device restriction is easier to justify when the offering genuinely cannot be delivered there or when the performance gap persists after the experience and measurement have been checked. Apply the smallest defensible restriction, then monitor whether volume shifts into more valuable inventory or simply disappears.

Steer channel delivery through assets, feeds, goals, and bids

Not every useful lever is an exclusion. In Performance Max, the material you supply tells the system what it can advertise, which formats it can assemble, which customers it should value, and what outcome bidding should pursue. These inputs influence delivery without offering an exact channel allocation switch.

Creative quality matters because Performance Max can serve across visual inventory including Display, YouTube, and Discover. Generic assets may technically make a campaign eligible for more formats while doing little to communicate the offer. Organize each asset group around one coherent product set, service, audience need, or landing-page promise. When several unrelated propositions share the same creative bundle, weak results become much harder to diagnose.

AI-generated images and videos can help fill missing formats and create variants, including assets derived from Shopping feed products. They still require human quality control. Before approving an AI asset, check:
- Whether the product, packaging, proportions, and important visual details remain accurate.
- Whether text is readable in the expected crop and does not introduce unsupported claims.
- Whether video motion, transitions, and product rendering remain coherent from beginning to end.
- Whether the message matches the destination page closely enough that the click does not create a new expectation.
- Whether the asset is acceptable for every type of inventory in which the campaign may use it.
The channel reporting view can show where delivery is occurring, but its actionable controls remain limited. If the campaign is appearing in a channel you would prefer to reduce, first inspect the inputs that made that inventory attractive: the asset mix, product feed, conversion goal, bid strategy, and budget. Changing these does not guarantee a particular distribution, but it addresses the logic the campaign is using.

When the business specifically needs Shopping-focused delivery, a feed-only campaign structure can concentrate the campaign on the product feed rather than supplying a complete cross-channel creative set. That choice trades broader creative reach for tighter inventory focus. Make it deliberately; do not remove assets simply because one channel report looks unfamiliar.

Conversion goals deserve the earliest inspection. If the campaign is rewarded for shallow actions that do not represent business value, exclusions will not solve the central problem. It will continue finding more of the outcome it was told to value. Make sure the selected goal represents a meaningful result and that different conversion actions are not being treated as equivalent when the business values them differently.

Bids and budgets are also steering mechanisms. They affect which opportunities the campaign can pursue and how aggressively it can compete, but they cannot repair an irrelevant goal or misleading creative. Fix the instruction before increasing the resources given to follow it.

Run the controls in a repeatable order

A control is useful only if you can connect it to a decision. Use one review sequence consistently so that urgent-looking reports do not pull you into random edits.
1. Record the current conversion goals, bid strategy, budget, schedule, exclusions, asset setup, and feed configuration. This is the baseline against which later changes will be judged.
2. Confirm that the campaign is optimizing for an outcome the business actually values. Resolve incomplete or misleading measurement before interpreting audience and inventory reports.
3. Review search terms. Add negatives only for clearly irrelevant or impossible intent, and record the reason for each important exclusion.
4. Review ‘Where ads have shown’. Use placement exclusions for documented suitability or relevance problems, remembering that an account-level action can affect other campaigns.
5. Inspect hour-by-hour delivery. Tighten the ad schedule only when the pattern is repeatable and consistent with the way the business handles demand.
6. Review demographic and device performance. Test whether the apparent gap comes from eligibility, the on-site experience, or measurement before removing reach.
7. Audit asset groups and feed inputs. Replace generic, inaccurate, mismatched, or low-utility material, and verify every AI-generated asset before it can represent the brand.
8. Use channel reporting to decide what to investigate. If strict Shopping focus is required, evaluate a feed-only structure; otherwise steer distribution through the available inputs.
9. Change one control layer at a time where practical. Annotate what changed, when it changed, and what outcome you expected.
10. Evaluate the next comparable period only after accounting for normal conversion lag. Keep changes that solve the stated problem; reverse those that merely reduce reach.
Start your next review with the search terms and placement reports, but do not stop at what looks wasteful. Trace each symptom back to the closest controllable cause. One well-supported negative, schedule adjustment, device fix, or asset correction is more useful than a dozen exclusions you cannot later explain.

References
- CrushPress.AI — Google PMax Enhances Control with New Exclusions & Reporting
- CrushPress.AI — Unlock Your Control Over Performance Max Campaigns
March 26, 2026

Tag: Budget Management

The three parts of a reliable Claude PPC system

Build the first workflow around one recurring decision

Put money-changing actions behind explicit gates

Choose PPC tasks by controllability, not novelty

Frequently asked questions

What is Claude-powered PPC automation?

Do you need MCP to use a Claude Skill for PPC?

Which PPC workflow should you automate first?

How do you prevent inconsistent Claude outputs?

References

Treat the manager account as hostile until you contain it

Recover each client account in a controlled order

Investigate the identity breach even when 2FA was enabled

Build an MCC that can fail without taking clients with it

Key takeaways for Google Ads MCC security and recovery

References

Start by proving that your ROAS means what you think

Build an account structure AI can learn from

Put guardrails around automated bidding

Prompt generative AI like an analyst with a proper brief

Turn every change into a controlled learning loop

Key takeaways

References

Treat conversion tracking as a bidding input, not a reporting detail

A practical validation sequence

Separate obvious waste from performance that needs more evidence

Reallocate budget instead of cutting every campaign evenly

Match bidding and creative decisions to the signal you actually have

For Shopping campaigns, product data is spend control

Key takeaways

References

Start with the conversion signal you actually want

Route each kind of intent to the right campaign treatment

Use precision where the intent justifies it

Let automation explore inside visible boundaries

Diagnose performance from revenue backward

Use AI to produce testable hypotheses, not synthetic certainty

Key takeaways

References

Replace the single funnel with a growth operating system

Build the growth brief before you write the agency brief

Divide ownership before you evaluate capabilities

Audit the agency’s decisions, not its pitch language

Ask six questions that require operational answers

Turn the winning answers into the working agreement

Key takeaways

References

Key takeaways: what to fix before spending more

Teach the buying system what a qualified lead means

Control the system through the inputs it learns from

Separate discovery from profitable scale

Use platform-specific structures without losing the common logic

Route Google Shopping demand through distinct layers

Organize Amazon Ads around the SKU and the commercial objective

Make product data, creative, and landing pages part of targeting

Run a human operating system around the automation

Key takeaways

References

Define efficiency before you buy another click

Build audiences from fit and intent, then keep them separate

Give each intent level its own campaign job

Make the ad qualify candidates before they click

Test the message against qualified outcomes

Use a two-stage application path and retarget real interest

Retarget people according to what they already did

Key takeaways for your next recruitment campaign

References

Key takeaways

Remove obvious waste with search terms and placement controls

Turn search-term visibility into a disciplined negative list

Use placement data for suitability before profitability

Apply time, demographic, and device limits without choking reach

Build an ad schedule around opportunity and operating capacity

Reserve demographic exclusions for durable mismatches

Diagnose the device experience before excluding the device

Steer channel delivery through assets, feeds, goals, and bids

Run the controls in a repeatable order

References