The headline number from BrightLocal's 2026 Local Consumer Review Survey is the one everybody quoted: 45% of consumers now use AI tools like ChatGPT to find local businesses, up from 6% the year before. That is a 7.5x jump in twelve months. It is the fastest documented shift in consumer search behavior in the last two decades.

The survey ran a representative panel of 1,002 US adult consumers via SurveyMonkey. That panel size is conventional for consumer research; the margin of error sits in the +/-3% range at standard 95% confidence. Below the headline, the data set contains roughly twenty other numbers that change how a local business should allocate marketing time and budget this year. Most of those numbers have not been quoted anywhere. This piece walks through them.

Putting 6% to 45% in context

Mobile search took about seven years to cross the 40% adoption mark with US consumers after the iPhone launched in 2007. Voice search took roughly five years to reach a comparable share of daily queries after the launch of Siri in 2011. AI-assisted local discovery did it in one. The closest historical analog is the early growth curve of social media, but social did not reach 45% of US adults until about three years after mainstream awareness of Facebook.

Our reading of this data suggests the compression has two drivers. First, AI consumer tools rode the existing distribution rails (mobile, browsers, app stores) instead of requiring new hardware. Second, ChatGPT crossed 800 million weekly active users by October 2025 (TechCrunch, October 2025), which gave the survey something concrete to measure against. A behavior change this fast does not usually slow back down to historical curves. It either keeps compounding or it plateaus inside two years. Either way, the strategic window for businesses to adapt is now, not when the curve flattens.

AI is the #3 local discovery source

BrightLocal's 2026 survey ranks AI tools as the third most-used source for finding local businesses, behind Google and Facebook (BrightLocal LCRS 2026). Google's share fell from 83% in 2025 to 71% in 2026.

The math here is worth sitting with. Google did not lose 12 points to AI alone. The consumer surface area is expanding: people use multiple tools for the same search, checking ChatGPT and then verifying on Google or Maps. But the trajectory of Google's decline matches almost exactly the curve of AI's rise on the same survey. If a marketing budget still allocates 80% to Google-centric channels (SEO, Google Ads, Google Business Profile optimization) and 0% to AI visibility, that allocation is calibrated to the 2024 consumer, not the 2026 one.

The trust numbers are more nuanced than the headline

Among active AI users, 63% trust AI platforms to provide local business recommendations and 64% trust AI as much as traditional review platforms. The numbers diverge sharply when you split the sample by whether the respondent actually uses AI (BrightLocal AI Trust report, 2026). Non-users default to skepticism.

The pattern is consistent with most platform adoption curves: trust climbs after use, not before. As the user base keeps expanding, the average trust level keeps climbing without any platform changing what it does. Today's active-user trust ceiling is tomorrow's general-population baseline.

Verification behavior: the AI-to-review loop

The most strategically useful number in the entire data set is this one: most AI users sometimes double-check AI recommendations against real reviews. A meaningful share check sources when reviewing AI recommendations. Almost half always verify against native review platforms.

What this means in practice: AI does not replace the review platform. It changes the job of the review platform. The new buyer journey looks like this:

  1. Consumer asks ChatGPT for a recommendation.
  2. ChatGPT returns 3-5 names.
  3. Consumer Googles each name and reads recent reviews.
  4. Consumer picks the one with the best review signal.

A business that is missing from step 2 never enters the funnel. A business that makes step 2 but has weak reviews loses at step 4. Both halves now matter at the same time. The old approach of optimizing reviews in isolation, or optimizing AI presence in isolation, leaves money on the table at the other step.

The 1.2% problem

Sitting next to BrightLocal's consumer-side data is a complementary supply-side number reported by National Law Review (March 2026): AI search platforms currently recommend only 1.2% of local business locations, and 83% of restaurants do not appear in AI-generated local recommendations at all.

Stack the BrightLocal and supply-side numbers on top of each other. 45% of consumers are asking AI. AI is only surfacing 1.2% of local businesses. The market is shifting consumer demand faster than the supply side is adapting. That asymmetry is the opportunity. If you fix the signals AI engines look for (structured data, directory consistency, review depth and recency, entity authority), you can move from the 98.8% to the 1.2% with a few months of work. The competitor next door almost certainly is not doing it yet.

Implications by business size

The right move depends on how many locations you operate. The data segments cleanly:

  • Single location or 1-5 locations. The asymmetric move. Most direct competitors are not optimizing for AI yet. A well-structured LocalBusiness schema, a Google Business Profile that matches across Yelp, Bing Places, and Apple Maps, and a steady cadence of recent reviews can lift you into the 1.2% inside a quarter. The cost is mostly time, not money.
  • Mid-size, 10-100 locations. The consistency play. The bottleneck at this scale is not strategy, it is execution. The same NAP (name, address, phone) has to be identical across 10-100 location pages, 10-100 GBP listings, and every major directory. One inconsistency per location compounds into hundreds of conflicting signals AI engines have to resolve, and they usually resolve by recommending a competitor with cleaner data.
  • Enterprise, 1000+ locations. The schema and entity graph play. AI engines do not just want correct information, they want machine-readable, hierarchically structured information that ties parent brand to child locations to services to reviews. This is a back-end engineering effort, not a marketing campaign. The ROI is large because the asymmetry is large.

What to do this quarter

Five concrete actions, each tied to a specific number in the data:

  1. Run an AI visibility audit on your business name in all four major engines. Tied to the 45% consumer adoption number. If you have never asked ChatGPT, Claude, Gemini, and Perplexity what they say about your category in your city, you are guessing. Most businesses score below 40 out of 100 the first time.
  2. Get to 20+ reviews if you are not there yet. Tied to the consumer-reviews threshold data. Below 20, you fail the screen for a large share of buyers regardless of how good your work is.
  3. Set up a review cadence that produces 4+ new reviews per month.Tied to the recency-of-reviews data. Stale profiles get filtered out at step 4 of the AI-to-review loop. A steady drip of recent reviews keeps you inside the consideration set.
  4. Add LocalBusiness and Service schema markup to every page that mentions a service or location. Tied to the 1.2% visibility number. AI engines reward businesses that hand them structured, machine-readable data. The other 98.8% mostly do not, which is what makes this asymmetric.
  5. Audit name, address, and phone consistency across Google, Yelp, Bing Places, Apple Maps, and your industry-specific directories. Tied to the review-platform verification step. AI engines triangulate. Inconsistent triangulation reads as low confidence and gets demoted.

A note on the methodology

BrightLocal's LCRS uses a SurveyMonkey representative panel of 1,002 US adult consumers, balanced across age bands. SurveyMonkey panels are a standard consumer-research instrument; they have known limitations around self-reporting bias (people may overstate AI use because it is a high-status behavior) and panel self-selection. Treat the absolute percentages as accurate within a few points and the year-over-year direction as highly reliable. The 6% to 45% trajectory is too large to be explained by methodology drift.

The takeaway

The 45% headline understates what is actually happening. Half of the active-AI-user group already trusts AI recommendations as much as reviews. Almost all of them cross-check against reviews before deciding. AI is currently surfacing only 1.2% of local businesses, while consumer demand has 7.5x'd in twelve months. The supply and demand curves are diverging fast, and the businesses that close that gap first will absorb the search volume that used to flow through Google rankings.

The fastest way to find out where you stand is to run a free AEO scan. It queries ChatGPT, Claude, Gemini, and Perplexity with 60 real prompts about your brand and scores you on brand recognition, competitive position, contextual relevance, sentiment, and citation authority. Sixty seconds. No account required. You will see exactly which of the four engines knows you exist and which are recommending your competitors instead.

Next steps once you have the data: read how to optimize for ChatGPT Search for the channel-specific playbook, and structured data for local businesses for the foundational schema work that lifts every engine at once.