Why 75% of ChatGPT Local Searches Still Use Keywords — What Local Businesses Must Do

Why 75% of ChatGPT Local Searches Still Use Keywords — What Local Businesses Must Do

New data reported by Search Engine Land shows that despite the rise of conversational AI, users searching for local services with ChatGPT often fall back to short keyword-style prompts. That finding has practical implications for local SEO and Generative Engine Optimization (GEO), and it should reshape how businesses prioritize search visibility across AI and traditional search engines.

What the data says

The Search Engine Land piece, reporting on an observational study, found that “across the observed sessions, 75% included at least one prompt that would be categorized as keyword-based,” a figure that contradicts the popular narrative that AI search is primarily conversational. The study focused on people using ChatGPT to find local healthcare and aesthetics providers and showed that many users prefer short queries such as “dentist 11214” or “best affordable dentist Manhattan.”

Additional details from the original research reinforce the point: 45% of sessions were one-shot prompts with no follow-up, and users typically considered multiple businesses (an average of 3.7) before deciding. The study’s authors also warned of attribution challenges when users discover a business through ChatGPT but later verify or click through using Google, which can mask the true origin of the lead in analytics.

Why keywords persist in AI-driven local search

There are three practical reasons keywords remain dominant for local service queries:

• Efficiency: Short keyword prompts require less effort and get users useful results quickly. Typing “dentist 11214” is easier than composing a long, detailed prompt.
• Satisficing behavior: Many users are satisfied with the first list provided and move straight to action, meaning long conversational interactions are unnecessary for transactional queries.
• Interface habits: People transfer long-standing search behaviors (from search engines) into new tools until those tools demonstrate clear benefits for more elaborate prompting.

As the Sagapixel study noted, “old habits die hard, and when your customers use AI to find local services, it looks a lot like Google keyword searches from 2003.” That reinforces the idea that GEO strategies should not abandon core keyword research.

Implications for local SEO and GEO

For local businesses and marketers, the headline takeaway is clear: continue investing in keyword-focused local SEO while also preparing for AI discovery. Specifically:

• Maintain and optimize classic local signals: Google Business Profile listings, consistent NAP (name, address, phone) data, local schema, and authoritative local citations remain critical to appear in both AI and traditional search results.
• Prioritize presence in the top consideration set: The average ChatGPT user examines several options. Appearing in the top 3–5 results — not just the top position — increases the chance of being chosen.
• Optimize for short transactional prompts: Ensure landing pages and local landing content reflect concise keyword phrases users are likely to enter (e.g., “dermatologist [zip]”, “botox near me”).
• Measure attribution differently: Expect some leads to be sourced via AI; update intake forms and call tracking to ask how customers found you, and use UTM best practices where possible to preserve attribution when users switch tabs.

Actionable steps for businesses

Here are specific tasks local businesses can implement this week:

1. Audit local listings: Confirm that Google Business Profile, Apple Maps, and major local directories have consistent NAP and category settings.
2. Map short-query keywords: Create a short list of high-intent, transactional local keywords (including zip-based or neighborhood terms) and ensure they appear in page titles, H1s, and local landing copy.
3. Build the consideration set: Produce content that increases the odds of being included in a multi-item answer set — concise service descriptions, transparent pricing or appointment info, and short FAQs that answer common transactional prompts.
4. Track offline conversions: Add a simple “How did you hear about us?” field to booking forms and train staff to ask during calls, so ChatGPT-originated leads are less likely to be misattributed.

Preparing for AI visibility beyond keywords

While keywords still matter, businesses should also prepare to be cited by AI systems in ways that go beyond simple on-page signals. Structured data (LocalBusiness schema), clear citations, and well-maintained knowledge panels will help LLMs and AI overviews find and trust your information. Equally important is building credible, citable content — reviews, local news coverage, and authoritative directory listings — that AI systems favor when assembling local recommendations.

As the Search Engine Land article emphasizes, these findings should prompt a measured approach: keep doing fundamental local SEO well, and layer GEO tactics that help AI find, verify, and cite your business accurately.

Attribution

This piece is based on reporting from Search Engine Land. Read the original article here: https://searchengineland.com/chatgpt-users-keywords-for-local-services-data-467715

Quote from the original article (Frank Olivo): “Across the observed sessions, 75% included at least one prompt that would be categorized as keyword-based.”

Additional context from the primary study (Sagapixel): “Old habits die hard, and when your customers use AI to find local services, it looks a lot like Google keyword searches from 2003.”

By SEOteric — https://www.seoteric.com