What actually changed?

GBP used to feed a map pack: humans compared listings and chose. In 2026 the same data feeds a synthesis layer. When someone asks Gemini or an AI Overview where to get their watch repaired or which dentist takes their insurance, the answer is assembled primarily from GBP fields, attributes, reviews, photos and Q&A — then cross-checked against your website and directories. Two consequences follow. First, completeness is no longer cosmetic: each empty field is a sub-question answered without you. Second, inconsistency became disqualifying — when GBP says one thing and your site another, the engine hedges or skips you. With the Gemini app past 750 million monthly users and AI Overviews past 2.5 billion (Alphabet, Q2 2026), this is the main channel for local demand, not an experiment.

Fields that now feed answers directly

  • Categories. The primary category is the routing key for best-X queries. Precision wins: Sichuan Restaurant beats Restaurant; Emergency Veterinarian beats Veterinarian — but only claim what is true, because reviews are cross-checked against claims.
  • Attributes. Languages spoken, accessibility, payment methods, service options — each is the answer to a real query family. Attributes you skip get answered from old reviews or not at all.
  • Hours, including holidays. Open-now queries are the highest-intent local queries that exist, and holiday hours (Lunar New Year, Thanksgiving) are the most-asked and most-wrong. Verify seasonally.
  • Q&A. Seeded and answered by you, this is answer-block content hosted on Google’s own citation source. Ten real questions, honestly answered, is an afternoon’s work with permanent returns.
  • Reviews. The synthesis material. Texture — named services, named staff, specific experiences — is what the engine quotes when it explains why it recommended you.

What about businesses that serve customers at their location?

Service-area businesses — plumbers, cleaners, mobile detailers — should hide the street address and define service areas precisely by city name, because engines answer which areas do you cover from that field plus your website text. Hybrid businesses (a salon that also does house calls) should configure both modes honestly rather than claiming a storefront they do not staff. Misconfigured service areas are a quiet killer: the engine either excludes you from near-me answers or, worse, sends customers to an address where nobody works.

The new maintenance cadence

  1. Weekly: respond to all new reviews; check for user-suggested edits (anyone can propose changes to your listing).
  2. Monthly: photos refresh, one or two posts, Q&A review, hours sanity check.
  3. Quarterly: full attribute audit against your actual service list; NAP sweep across directories.
  4. Seasonally: holiday hours verified before every major holiday window.

Where GBP fits in the wider graph

GBP is necessary, not sufficient. ChatGPT’s local layer leans on Bing — so claim Bing Places and submit IndexNow. Your website must corroborate GBP facts with matching text and LocalBusiness schema. Directories must carry identical NAP. The profile is the keystone, but the arch is everything around it. Full pipeline: the GBP-to-AI visibility playbook and the local business GEO hub.

The businesses AI recommends are not the ones with the cleverest optimization. They are the ones whose facts are complete, consistent and corroborated everywhere the engine looks. GBP is where that discipline starts.

Frequently asked questions

Is GBP still worth it if I rank well organically?
More than before. Organic rank feeds the retrieval pool, but GBP feeds the answer itself — hours, attributes, reviews. A strong site with a neglected profile loses local AI answers to a complete-profile competitor with a modest site.
How do I know if AI answers are using my GBP?
Ask the engines your category questions monthly and screenshot the answers: the facts quoted (hours, attributes, review phrases) trace directly to GBP fields. The four-dimension measurement loop formalizes this check.