I’ve got some exciting news about Google Ads: They’ve introduced something called App Consent Insights! This new feature aims to give us, the advertisers, a much clearer picture of how consent affects our app campaign performance.
What’s new? There’s this cool diagnostics view that breaks down consent data across various apps, platforms, regions, and traffic sources. It’s a game changer for understanding where we might have gaps in our setup.
Zoom in. I can now see an overall consent rating described as “Excellent,” “Good,” or “Poor.” Plus, there’s a live count of apps actively sending consented data and a detailed table that shows consent rates for conversions, including the differences between EEA and non-EEA users.
Why it matters to us. With privacy regulations getting stricter, consent isn’t just a compliance issue—it’s a critical factor for measurement and optimization. This update gives us more visibility into how consent setups could be holding back our performance.
Between the lines. Google is making it easier for us to measure and act on consent data at a time when signal loss significantly impacts campaign performance.
What to watch. We should start looking at optimizing not just for conversions, but also for improving consent rates as another lever of performance.
Bottom line. With better visibility into consent, we can achieve better data quality and ultimately, better campaign outcomes.
First seen. Google Ads expert Thomas Eccel first noticed this update on LinkedIn.
I’ve been noticing the rapid transformation in how brands are tracking user behavior online. With privacy laws tightening and browser extensions increasingly blocking data, the demand for cleaner data from ad platforms is higher than ever. This change urged me to explore server-side tagging as a solution.
By implementing server-side tagging, I’ve managed to reduce data loss while collecting cleaner, privacy-compliant data. This approach is invaluable, especially considering the experiences I’ve had with providers like Elevar and Littledata.
So, what exactly is server-side tagging, and in which situations does it really shine? Let’s dive into the details!
What is server-side tagging?
Traditionally, tracking scripts ran directly in the browser. However, with server-side tagging, these scripts operate on a server I control, giving me more control over data processing.
Here’s how it works: instead of sending data straight to multiple third parties from the browser, events are sent to a first-party server endpoint, often using a Google Tag Manager server-side container. The server then processes, enriches, and forwards this data to tools like Meta and Google Analytics.
This setup provides benefits such as more data control, a cleaner page performance, and better compliance with privacy laws.
Moreover, server-side tagging grants me the flexibility to enrich and transform data before it reaches ad platforms, standardizing event names, filtering out low-quality events, and adding custom parameters for better audience segmentation.
Is server-side tagging right for you?
While server-side tagging isn’t a one-size-fits-all solution, many brands find it essential, particularly if you:
You need to meet strict privacy or compliance requirements
Server-side setups allow for greater control over how data is processed and shared, supporting compliance with regulations like GDPR and CCPA.
You want faster website performance
In my experience, client-side tracking can slow your page down, but server-side tagging shifts data processing to the server, resulting in faster websites.
You want more accurate tracking (despite ad blockers)
Ad blockers can hinder client-side scripts, but server-side tagging circumvents many of these restrictions, making your data collection more reliable.
You’re investing heavily in paid media
For those heavily invested in platforms like Meta and Google Ads, achieving better data accuracy can significantly impact return on ad spend.
How to implement server-side tagging
When it comes to implementing server-side tagging, you have two main options: building it internally or using a service provider.
Option 1: Internal setup
Choosing an internal setup gives me complete control but requires technical expertise and ongoing maintenance. This involves setting up a GTM server-side container and adding logic for data processing.
Option 2: Use a server-side tagging service
Platforms like Elevar and Littledata offer turnkey solutions that integrate seamlessly with existing tools, allowing me to focus on strategy rather than technicalities.
Our direct experience: Littledata vs. Elevar
In my experience with Littledata and Elevar, each caters to different needs. Littledata is ideal for emerging brands with simpler tech stacks, while Elevar is suitable for those outgrowing entry-level solutions.
Investing in server-side tagging has transformed how I handle data, ensuring that I remain compliant with privacy laws while boosting site performance and data reliability across all my platforms.
I’ve been following the latest updates from OpenAI, and they recently made some significant changes to their privacy policy, especially with the introduction of ads in ChatGPT. These updates are designed to allow advertisers to run personalized ads while ensuring that our chats remain private and secure.
OpenAI shared these updates with ChatGPT users, detailing how ads will function within the platform and clarifying what data is accessible to advertisers. It’s a refreshing assurance that our personal interactions remain confidential.
Why this matters to me. Privacy is paramount, and OpenAI emphasizes that personal chats and histories remain shielded from advertisers. They utilize anonymized engagement signals for ad personalization, ensuring advertisers can target relevant users without accessing sensitive information.
This method allows advertisers to evaluate the performance of their ads within a privacy-first framework, fostering user trust.
Ads in ChatGPT For users like me on Free and Go plans, ads might start appearing, but if you opt for paid tiers like Plus, Pro, Enterprise, Business, and Education, you can enjoy an ad-free experience. OpenAI promises clear labeling and separation of ads from chatbot responses.
Importantly, the content generated by ChatGPT remains unbiased and unaffected by these advertisements.
How ad targeting is handled. OpenAI uses in-platform signals such as ad interactions to personalize ads, but advertisers do not get access to our conversations, chat histories, or personal information.
Advertisers receive only aggregated metrics like total views or clicks, ensuring our personal data stays protected.
Additional privacy updates A new feature allows for optional contact syncing, helping us connect with friends who also use OpenAI services. It’s up to us whether to enable this feature.
They also provided more transparency on data storage durations, processing methods, and user control options, helping us understand our data management better.
Safety and product enhancements. The update encompasses new safety tools and age prediction systems aimed at ensuring a safer environment for teenagers. Documentation for new features like Atlas, Sora 2, and parental controls for teen accounts has also been included.
The bottom line. With the expansion of advertising in ChatGPT, OpenAI is committed to maintaining strict boundaries concerning user privacy, offering advertisers valuable insights without infringing on personal conversations or data.
This update was first spotted by Paid Media expert Arpan Banerjee, who shared insights on LinkedIn. It’s a promising move towards privacy-centric advertising in AI-powered platforms.
On the OpenAI podcast, I recently listened to Andrew Maine as he spoke with OpenAI executive Assad Awan. During their conversation, Awan shared insights into how ads are being introduced to ChatGPT, who will see them, and the measures in place to protect user trust.
Who will see ads:
Ads will be visible to users on the Free and Go tiers. As for Plus, Pro, and Enterprise subscribers, they will not encounter ads in their interactions. Additionally, Enterprise workspaces are staying completely free from advertisements.
The guardrails: Awan highlighted that OpenAI is committed to structuring ads with strict trust principles in mind.
Separation: Ads are distinctly separate both visually and technically from the model answers.
Privacy: Conversations are not shared with advertisers, ensuring privacy is upheld.
Sensitive topics: Discussions on health, politics, and other sensitive subjects will never be interrupted by ads.
Controls: Users have the ability to adjust ad personalization settings or even upgrade to remove ads entirely.
Awan also mentioned that the AI model itself is not aware of when ads are present and will only reference them if directly queried by a user.
Zoom in. OpenAI emphasizes prioritizing user trust over other factors such as user value, advertiser value, and revenue. This framework is designed to prevent ads from influencing the model’s responses.
For small businesses. Awan envisions a future where AI simplifies advertising for small businesses. By understanding plain language goals, AI can help run campaigns without the complexity of traditional dashboards.
Why we care. ChatGPT ads promise a unique, high-intent channel where businesses can connect with users during their active conversations and decision-making processes. By focusing on relevance and AI-driven matching, the platform can lower the entry barrier for small to midsize advertisers while boosting performance for larger brands.
Should OpenAI succeed in cultivating a trusted ad environment, it could reshape how advertisers perceive discovery and customer engagement within AI-driven platforms.
What’s next. The initial ad tests will remain conservative, concentrating on utility and relevance before volume as OpenAI hones ad formats and placements.
The big picture. Through advertising, OpenAI aims to expand ChatGPT access while adhering to a trust-first design—a balance they assert is key to their long-term strategy.
I often find myself reflecting on the challenges of PPC measurement in this privacy-driven era. As browser restrictions tighten, our reliance has shifted from perfect tracking to methods like redundancy, modeling, and inference.
Managing PPC accounts has shown me firsthand that something has changed. The signs are everywhere:
Missing GCLIDs, delayed conversions, and reports that are harder to explain have become routine.
Initially, it feels like something broke—perhaps a tracking update or a platform shift. Yet, it’s simpler than that. We often assume identifiers will persist from click to conversion, but that’s no longer a reliable expectation.
Measurement hasn’t ceased to function; what has changed are the conditions it relies on. These changes have been creeping up, gradually becoming the norm.
Why this shift feels so disorienting
Having dealt with this issue for most of my career, I find the evolution quite disorienting. Before native conversion tracking in Google Ads, I crafted my tracking pixels and parameters for affiliate campaigns. Moving towards automation and less control can feel unsettling compared to the traditional methods.
The things I once depended upon for PPC data interpretation don’t apply in the same way anymore. Adjusting my mindset is key to navigating this evolved landscape, as restoring the old assumptions won’t work.
The old world: click IDs and deterministic matching
Predictability was the hallmark of Google Ads measurement. A click led to a gclid being stored in a cookie and matched back upon conversion, creating an easy-to-explain deterministic system.
This depended heavily on things like parameters passing through browsers and cookies persisting. Thankfully, these conditions were favorable back then.
Why that model breaks more often now
Today’s browsers impose stricter limitations on identifiers. Apple’s Intelligent Tracking Prevention and similar techniques significantly reduce tracking data’s shelf life, directly impacting how data is stored, or even if it can be stored.
On occasions, click IDs fail to reach the site, and the behavior of browsers today necessitates adapting, rather than attempting to cling to outdated deterministic systems.
The adjustment isn’t just technical
On my team, GA4 poses challenges not because it’s ineffective, but because it suits a reality where some data is presumably missing. This experience is shared industry-wide; we must acknowledge that measurement now requires a new mentality.
Ultimately, surviving in this privacy-centric era demands refocusing energy on resolving data problems rather than merely optimizing ad settings.
What still works: Client-side and server-side approaches
The question now is how we can thrive under current constraints, and the answer involves both client-side and server-side measurement practices.
Pixels still matter, but they have limits
Though these pixels provide valuable data and instant feedback, their efficacy is limited by browser constraints and consent banners blocking data.
Relying solely on pixels for measurement affects both our reporting and optimization efforts. While they’re not obsolete, they no longer cover every base.
Changing how pixels are delivered
In search of better solutions, some focus on improving pixel delivery, such as Google Tag Gateway, which routes tags through the same-origin setup. This minimizes failures but does not define better measurement logic by itself.
There’s a distinction between improved infrastructure and improved measurement logic—we must remember that proper data collection and event definition are crucial.
Offline conversion imports: Moving measurement off the browser
Using offline conversion imports moves measurement away from browsers to backend systems, mitigating browser-imposed privacy restrictions and extending its efficacy to longer sales cycles.
Combining offline imports with pixel tracking ensures a complete view of customer interactions.
Even without click IDs, Google Ads utilizes other inputs to match conversions, although we must be aware that modeled data fills gaps when consent is denied or IDs are missing.
Even with complete information from pixels or offline imports, conversions sometimes remain elusive.
Determining how this aligns with privacy restrictions and user consent requires ongoing refinement and a strategic approach.
Designing for partial data
Partial data is now the status quo, and including multiple sources of input can create a robust strategy to overcome discrepancies in systems like CRMs and Google Ads.
By learning to accept this tension and strategically managing incomplete data, we can optimize campaigns for the current data landscape.
As we embrace a privacy-focused measurement strategy, perfect tracking is no longer feasible. Building useful measurement systems requires recognizing differing operational views and aligning accordingly.
Ultimately, strategic thinking, redundant data systems, and careful evaluation are essential components in adapting to a partially observable data world.
Today’s measurement landscape demands a strategic approach instead of recreating past perfection.
I recently stumbled upon a fascinating preview of ChatGPT’s new ad configurations, giving us an insight into how personalization and privacy will revolutionize ad delivery within conversational AI.
Driving the news. It was an exciting moment when Juozas Kaziukėnas, an innovative entrepreneur, uncovered a method to access ChatGPT’s forthcoming ad settings interface. The panel is reassuring in its consistent emphasis: advertisers won’t have access to our chats, history, personal details, or IP addresses.
What the settings reveal:
There’s a well-organized ad framework complete with its own controls.
A History tab, where I can check the ads I’ve viewed inside ChatGPT.
An Interests tab that gathers inferred preferences based on my interactions and feedback.
For each ad, I have the option either to hide it or report it.
Importantly, I can delete my ad history and interests without affecting other ChatGPT data.
Personalization options. I have the freedom to turn ad personalization on or off. When it’s enabled, ChatGPT uses my saved ad history and interest cues to customize ads. If disabled, the ads still display but only consider my current conversation for relevance.
An intriguing option allows ad personalization using both past conversations and memory capabilities — though crucially, my chat content isn’t shared with advertisers. For accounts like mine with memory disabled, this feature remains inactive.
Why we care. Even though official ads haven’t launched, the newly accessed settings panel provides us with the most detailed preview yet of ad personalization and privacy controls in action. It’s exciting to see ChatGPT striving to balance effective personalization with rigorous privacy standards. I can already imagine how this will redefine ad targeting and measurement on the platform.
The settings indicate a focus on contextual signals and user-enabled personalization, avoiding overly intrusive user tracking. This means our creative relevance and the intent derived from our conversations will be valued more than conventional audience profiling.
For brands, it’s a hint on how to craft their messaging and strategies for this new wave of conversational advertising.
The bigger picture. This discovery suggests OpenAI is developing an ad system mirroring known platforms but with a fresh focus on privacy and user autonomy.
Bottom line. Although ChatGPT ads might not be live right now, the framework is clear and indicates a future where conversational ads offer nuanced privacy and personalization settings.
First seen. Kaziukėnas shared a preview of the platform on LinkedIn.
I’ve noticed that OpenAI is introducing premium-priced ads on ChatGPT, but here’s something interesting: the data provided to advertisers is significantly limited compared to what we’re used to.
What’s happening. Reports indicate that OpenAI is offering ChatGPT ads at around $60 per 1,000 impressions. That’s about three times the rate of standard Meta advertisements! Yet, even with this higher cost, advertisers only receive basic metrics like total impressions or clicks, without insight into actions like purchases.
Why we care. ChatGPT is becoming a fresh, highly engaging ad space, but it’s not without its challenges. The hefty CPMs and limited insights mean that early advertising efforts will lean more toward enhancing brand presence and gathering learnings than achieving performance-driven efficiency.
For marketers who are open to trying new avenues, this presents a unique chance to gain insights into how ads function within AI-driven conversations before the format becomes more widespread or measurable.
The tradeoff. OpenAI is contemplating expanding its measurement capabilities in the future, yet it remains committed to user privacy. It has pledged not to sell user data or invade the confidentiality of conversations, which limits traditional targeting and attribution possibilities that platforms like Google and Meta offer.
Who will see ads. Initially, these ads will be available to those using ChatGPT’s free and lower-cost Go tiers, but won’t be shown to users under 18 or in conversations concerning sensitive topics like mental health or politics.
Between the lines. OpenAI is branding ChatGPT ads as a top-tier, trustworthy product, banking on the idea that context, focus, and brand safety can validate the higher pricing, despite the lack of detailed performance data.
Bottom line. Brands eager for prominent visibility in a cutting-edge AI-driven environment may find ChatGPT ads appealing, but those focused on performance metrics might hesitate due to the absence of detailed measurement.
As I delve into the recent statements from Google, I am struck by the urgency in Elizabeth Reid’s affidavit. She warns us that if Google is compelled by the court to share its search index and ranking data, it could seriously jeopardize user privacy, potentially inviting spam abuse.
Reid, who heads Google’s Search department, presented her affidavit as part of Google’s motion to pause the implementation of some antitrust remedies. Her warning highlights the potential “immediate and irreparable harm” that such data sharing could cause to both Google and its users.
What strikes me is how Reid articulates the danger of exposing Google’s sensitive Search assets, which could lead to reverse engineering and an escalation in spam.
Imagine, for a moment, how revealing the web search index could become problematic. Under the court’s Section IV ruling, Google might have to provide competitors with crucial web index data. This includes every URL in Google’s index, a DocID-to-URL map, and more. For us at Google, this just seems like handing over the results of 25 years of meticulous work.
Reid explains that the web index is born from proprietary systems that decide the inclusion of pages in Google Search. Knowing which URLs are indexed by Google could allow potential competitors to bypass comprehensive crawling, thereby gaining undue advantage.
Further adding to the complexity, metadata like crawl frequency offers insight into how Google prioritizes content, which again, could provide competitors with unfair advantages if unveiled.
Reid’s affidavit includes images illustrating Google’s processes. One notably shows most webpages labeled as “Spam, Duplicates, & Low Quality Pages,” an insight into how meticulous our web crawling is. It’s fascinating to think that as of 2020, Google’s index boasted around 400 billion documents.
There is also a dire warning about exposing spam scores. Such a leak could greatly weaken Google’s spam-fighting mechanisms, making it harder to protect users from low-quality content.
In terms of user data, the transparency required by the judgment would mean sharing extensive search logs used by Google’s Glue and RankEmbed models, including detailed user interactions. This suggests a large-scale disclosure of Google’s proprietary data signals, something Reid is quite concerned about.
Finally, the requirement to syndicate Google’s core search results to competitors for five years poses a significant challenge. Despite contractual limits, our control over our systems would diminish, with possible data misuse or leaks.
Reid’s testimony underscores her knowledge and dedication as she stands by Google’s motion to stay antitrust remedies while the appeal is pending. If you’re interested, you can explore Reid’s affidavit further.
Recently, I discovered that Google is offering advertisers more control over data flow, which is especially helpful when user consent is limited.
Driving the news. There’s a new tool out called Data Transmission Control, appearing in Google Ads. This enhancement builds on Advanced Consent Mode by providing a more detailed approach to managing how advertising, analytics, and diagnostic data are shared.
What’s new. As an advertiser, I can now independently adjust the flow of advertising data, behavioral analytics, and diagnostic data. If ad_storage consent is not given, I have two choices: either allow limited data with identifiers removed (which still supports conversion modeling), or entirely block the data until consent is obtained. Interestingly, I can still allow behavioral analytics even if ad data is restricted, or choose to block it completely.
Where to find it. I found the setting hidden within Data Manager → Google Tag (Manage) → Manage data transmission. It’s easy to overlook if you’re not looking carefully.
Why we care. Traditionally, Consent Mode was all about reflecting user choices. Now, with Data Transmission Control, I can decide—right down to the tag level—what data flows when there’s no consent, aligning more closely with privacy-focused strategies.
It’s empowering to have this degree of control, especially when trying to balance privacy compliance against performance metrics, which is crucial in markets with strict regulations.
Key details. It’s important to note that Consent Mode must be enabled for this feature to function. It’s set up via the user interface in Google Ads, Google Analytics, or Campaign Manager 360, and applies only to Google tags. If the feature isn’t enabled, everything stays the same, but once consent is given, data transmission resumes automatically.
First seen. This update was first reported by Google Ads expert Thomas Eccel, who shared his insights on LinkedIn.
The bottom line. The introduction of Data Transmission Control provides a subtle yet powerful way for me to ensure tighter data collection control without fully losing out on valuable measurement capabilities.
Google just introduced a beta integration for the Google Tag Gateway, allowing advertisers, like myself, to deploy it effortlessly through the Google Cloud Platform (GCP). The process is now simplified with a new one-click workflow available in Google Tag Manager and Google tag settings.
What’s really exciting is how the GCP integration leverages Google Cloud’s Global External Application Load Balancer. This tool routes tag traffic through our own first-party domain before sending it off to Google, which enhances the deployment process. This strategic approach not only improves data signal quality but also boosts resilience against ad blockers and features like Apple’s Intelligent Tracking Prevention.
Why does this matter to us? As third-party tracking faces increasing limitations from browsers and platforms, advertisers like us need reliable ways to protect measurement signals. By directing Google tags through our infrastructure, we can maintain the integrity of our measurement signals against ad blockers and browser privacy constraints.
For those of us already using Google Cloud, this one-click setup significantly reduces the barriers to achieving more resilient and future-proof tracking.
What are others saying? Digital marketer and Simmer co-founder Simo Ahava highlighted this advancement on LinkedIn. According to him, the integration facilitates a seamless GCP deployment. It automatically configures an External Application Load Balancer with rules to direct Google Tag Gateway traffic to our backend services handling these requests.
Ahava also noted that Google Tag Gateway positions Google’s tagging infrastructure behind a same-site, same-origin first-party host, ensuring that tags endure in restrictive browser environments.
The broader perspective here is that previously, Cloudflare was the only automated option for deploying Google Tag Gateway, with other CDNs requiring manual setups. By adding GCP, Google reduces the friction for us advertisers already committed to their cloud ecosystem, thus promoting first-party tagging strategies.
The bottom line? Google is simplifying first-party tagging deployment, and while the GCP integration is still in its beta stage, it represents a significant stride toward robust measurement solutions in our increasingly privacy-focused digital landscape.
