75% of ChatGPT Users Still Prefer Keywords for Local Services

```json
{
  "alt": "Search bar with the query 'dentist in Chicago' entered.",
  "caption": "Looking for dental care in the Windy City? The search bar suggests finding a dentist in Chicago quickly and efficiently.",
  "description": "This image showcases a digital search bar with the query 'dentist in Chicago' typed in, accompanied by options to attach, search, or study. The interface suggests a user-friendly way to find dental services within the Chicago area, highlighting the convenience of online search tools. Keywords include dentist, Chicago, search bar, and online search."
}
```

75% of ChatGPT users rely on ‘keywords’ for local services- New data

New data has revealed that many people, like myself, use ChatGPT to search for local healthcare and aesthetic services using short prompts and specific keywords.

75% of ChatGPT users rely on ‘keywords’ for local services- New data

In the SEO world, there’s been this idea that people have changed how they search for local services, preferring longer, conversational prompts over simple keyword searches.

With tools like ChatGPT becoming more prevalent, I wanted to see how true this is. I observed how everyday users, including myself, used ChatGPT for locating local service providers like healthcare and aesthetic practices.

We all began our search with ChatGPT, mimicking our usual behavior—be it visiting websites, checking social media profiles, or reading reviews.

Some key questions guided our observations:

  • Are we using ChatGPT the way industry experts assume, especially when it comes to searching for local services?
  • Is the trend moving away from keyword searches, rendering traditional keyword strategies obsolete?
  • Do people truly engage in extended dialogues with ChatGPT for transactional purposes?

The results were intriguing, as they debunked many widespread beliefs being touted in SEO circles.

75% of Sessions Included Keyword Searches

Despite assumptions to the contrary, it turns out I’m not alone in still relying heavily on keyword searches when using AI platforms.

```json
{
  "alt": "Search bar with query 'dentists montgomery county pa' displayed.",
  "caption": "A search engine inquiry seeking dental services in Montgomery County, PA, highlighting the importance of accessible healthcare information online.",
  "description": "This image shows a search bar with the query 'dentists montgomery county pa'. It's a representation of users looking for dental healthcare services in Montgomery County, Pennsylvania. The visual indicates a fundamental online search, emphasizing the need for easy access to healthcare listings and information. Keywords: search query, dental services, Montgomery County, Pennsylvania, healthcare information."
}
```

Initially, we didn’t focus on keywords, influenced by the misconception that they were becoming obsolete in AI contexts. However, observing user behavior revealed a pattern very familiar from traditional search engines.

To my surprise, 75% of the sessions I observed included at least one keyword-based prompt, reaffirming the continued relevance of keyword-based search behavior.

Here are a few screenshots showcasing this behavior:

ChatGPT - dentist in chicago
ChatGPT - dentists montgomery

Honestly, this should not come as a surprise.

It’s much easier to type a short phrase like “dentist 11214” than to input a long sentence explaining every detail.

This behavior aligns with how users have traditionally interacted with search engines—a habit that’s hard to break.

These observations raise questions about the relevance of keyword tracking in the evolving landscape of Generative Engine Optimization (GEO).

```json
{
  "alt": "Search input field displaying 'botox by a doctor philadelphia' with misspellings underlined.",
  "caption": "A search input field suggests readiness for a query on Botox treatments in Philadelphia, highlighting some spelling corrections.",
  "description": "This image shows a search input field with the text 'botox by a doctor philadelphia.' Words 'botox' and 'philadelphia' are underlined, indicating spelling suggestions. The phrase 'Ready when you are' appears above, suggesting the interface is ready for a search. This image highlights user interaction with search functionality, focusing on Botox procedures in Philadelphia."
}
```

Some propose that GEO should transform transactional keywords into elaborate sentences, yet for local services, this seems unnecessary.

Despite varied responses from AI learning models, keywords still often appear when users seek services, affirming their utility in GEO strategies.

Dig deeper: Why local SEO is thriving in the AI-first search era


Local Searches Aren’t Conversational: 45% Use One-shot Prompts

When I’m searching for local services, I’ve found that engaging in a lengthy dialogue with ChatGPT is quite rare.

In fact, nearly half of the sessions—45%—were resolved with a single prompt, without any additional questions.

Additionally, when follow-up prompts were used, 34% simply requested more results.

At an average, I would say most users employed around 2.1 prompts during their searches for local services.

TaskAverage number of prompts
Find a new dentist2.41
Find a place to get botox1.96
Find a dermatologist to check a mole1.71
Hair transplant1.33
Find a chiropractor2.33
Decided to get a facelift2.00
```json
{
  "alt": "Bar chart showing number of prompts used in ChatGPT for local searches, measured against sessions.",
  "caption": "Exploring how many prompts users typically give ChatGPT for local search inquiries. An insightful study into AI-assisted local searches.",
  "description": "This image features a bar chart depicting the number of prompts used in ChatGPT sessions specifically for local searches. The x-axis represents the number of prompts, while the y-axis indicates the number of sessions. The chart, conducted by Sagapixel, highlights user interaction patterns with AI in local search contexts. This visualization is essential for understanding user engagement levels with AI tools in localized scenarios."
}
```

This doesn’t mean conversations aren’t happening; instead, the notion of conversational searches for local services appears overestimated.

Dig deeper: GEO x local SEO: Future of Discovery

Prompting Varies by Intent & Could Evolve

It seems people engage in conversations with ChatGPT mostly for informational needs, differing from what my findings might suggest.

However, documenting this behavior proved challenging due to its support largely from anecdotal evidence.

When it comes to services, we’re still more inclined to prompt ChatGPT as if we were using Google or another search engine.

This could change as users become more familiar with upgraded, paid versions of ChatGPT.

Yet, the simplicity of keyword searches may always hold appeal, especially if AI models efficiently deliver desired results with minimal input.

Explore the complete study and data for more insights. (Disclosure: I’m the founder of Sagapixel.)

Dig deeper: Boost Local SEO: 4 AI-Driven Tactics


Inspired by this post on Search Engine Land.


crushpress.ai community screenshot

FAQs

What percentage of sessions included keyword searches?

75% of sessions included keyword-based prompts, underscoring the continued relevance of keywords in local service searches. This finding challenges the notion that keyword searches are obsolete in AI contexts.

Are local searches conversational according to the study?

45% of sessions were resolved with a single prompt, and 34% used follow-up prompts to request more results. On average, users submitted about 2.1 prompts per search, indicating that local queries are not highly conversational.

Does the post challenge the idea that keyword searches are obsolete in AI contexts?

Yes. The post notes that keyword-based searches remain common when using AI platforms, challenging the assumption that keywords are obsolete, and it questions the relevance of keyword tracking in GEO.

What resources are linked for deeper insights?

The post links to three resources for deeper insights: ‘Why local SEO is thriving in the AI-first search era’, ‘GEO x local SEO: Future of Discovery’, and ‘Boost Local SEO: 4 AI-Driven Tactics’.

Who authored the post?

The post is authored by shivamcrushpressai. It also includes a disclosure stating the author is the founder of Sagapixel.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *