What does the local answer layer look like in late 2026?
The local discovery journey has compressed into a single synthesized response. A consumer asks where should I take my car, which dentist takes my insurance, best dim sum near me — and receives two or three names with reasons attached: open now, 4.7 stars with reviewers mentioning honest pricing, specializes in European cars. Scale anchors: Google AI Overviews above 2.5 billion monthly users (Alphabet, Q2 2026) and the Gemini app above 750 million monthly actives, with local answers assembled primarily from GBP, Maps and the review corpus. ChatGPT (~900 million weekly users) handles a growing share of local queries through its Bing-dependent retrieval, which is why Bing Places quietly matters. Perplexity passes a billion monthly queries with sentence-level citation of local sources. The map pack did not disappear; it was absorbed into the answer.
Which signals decide local recommendations
Across the local verticals we analyze, the same five inputs recur in the engines’ stated reasons:
- GBP completeness and precision — categories, attributes, hours (especially holiday hours), seeded Q&A.
- Review texture — not the star average alone, but the quotable specifics: services named, staff named, situations described. Engines quote phrases, not averages.
- NAP consistency — entity confidence across directories; contradictions read as risk.
- Website answer content — pricing shape, insurance lists, service areas, policies, in text with LocalBusiness and FAQ schema.
- Freshness — recent reviews, current photos, updated hours; stale profiles lose to active ones in time-sensitive queries.
Which verticals are moving fastest?
Based on the query patterns engines expose and the verticals covered in our playbooks: restaurants and home services moved first (high query volume, deep review corpora), healthcare-adjacent categories follow with more hedging under YMYL scrutiny, and considered-purchase verticals — legal, accounting, real estate — are where question-shaped content creates the largest openings because competitors publish the least. Across all of them the pattern is identical: complete facts plus review texture plus quotable answers beats adjective-rich marketing, every month we measure.
The zero-click reality for local
Local intent was zero-click before it was fashionable — calls and direction requests never touched websites. AI answers deepen the pattern: the Ahrefs February 2026 study (300,000 keywords) found position-one CTR down roughly 58 percent under AI Overviews and about 68 percent of searches ending without a click. For local businesses the practical meaning is unchanged but sharper: the answer is the storefront. Being named in it is the metric; website sessions are a side effect.
What local businesses should measure
A fixed set of local questions — best-X, open-now, takes-my-insurance, cost-of-service for your metro — run monthly across Google, ChatGPT and Perplexity, scored on the four-dimension standard with screenshots archived. Visibility and Brand Perception move first (weeks, as GBP and reviews improve); Citation Rate follows as website answer content matures (a quarter); Semantic Coverage is the long game. Vertical playbooks with the specifics live in the local business GEO section.
The honest caveat
Local AI answers remain volatile: engines rotate sources, personalization varies, and a single bad month of reviews can move answers faster than a good quarter of content. Measure trends over two to three months, never single runs — and treat any vendor promising a fixed local AI ranking the way you would treat a guaranteed-SEO salesman from 2010.