Google Business Profile in 2026: The 11 Settings That Still Move the Dial

Google Business Profile is the single highest-leverage free surface a restaurant owns. It is also the surface most operators treat as a "set it and forget it" task — completed once in week one of the venue's life, then forgotten until something breaks.

In 2026 that approach is no longer survivable. The profile is no longer just a Maps pin. It is the structured input that Google's AI Overview, ChatGPT browsing, Perplexity, Apple Intelligence and every other emerging answer engine read to compose responses about your venue. Settings that were "nice to have" in 2022 are now ranking factors. Settings that were ranking factors in 2022 are now thresholds.

Here are the eleven we revisit weekly on every Kitxens-managed profile, in priority order by their effect on discovery.

#1 — Primary category

The single most important setting on the entire profile. It must match the exact phrase diners use to describe what you are.

Common error: A modern Italian trattoria in central Madrid listed as "Restaurant" instead of "Italian restaurant" — losing roughly 60% of relevant local queries because the model never associates the venue with the cuisine.

Audit move: search for your venue type from a private browser window, in the geography you serve. If your venue does not appear in the first ten results for the most common phrasing, your primary category is wrong.

#2 — Secondary categories (up to nine)

Underutilized by 80% of independent operators. The model uses these to compose nuanced answers — "Italian restaurant with vegetarian options that takes reservations" — even when each individual category alone would not have surfaced you.

We typically use seven of the nine slots. Examples for a typical mid-market modern Italian: Italian restaurant, Pizza restaurant, Pasta restaurant, Wine bar, Catering service, Lunch restaurant, Dinner restaurant.

#3 — Service area and dining options

Explicit, not implied. "Dine-in: yes. Takeaway: yes. Delivery: yes. Curbside pickup: yes." If a setting is true, set it. If it is false, set it explicitly to no. Empty fields are read by the model as "unknown" and downgrade citation likelihood.

#4 — Menu link

Direct to your menu page, not your homepage. The model wants to read menu items to compose answers about dietary fit, price band and cuisine specifics. Linking to the homepage and burying the menu behind a click is a meaningful citation tax we observe consistently in our audits.

If you have a multi-location group, link each location's profile to the location-specific menu — not the brand-wide menu. Granularity is rewarded.

#5 — Reservation link

Connected to your booking provider (OpenTable, The Fork, SevenRooms, Resy, etc.), not "call us." The AI layer increasingly filters by "takes reservations online" — and a phone-only venue is now invisible to those queries.

If you cannot afford a third-party reservation provider, a free direct-booking form on your own site, linked here, is still meaningfully better than a phone number.

#6 — Order link

To your direct ordering site, not the marketplace. This is the single most under-set field in our audit data. Operators are spending money to drive traffic to Google, then handing the traffic to Uber Eats by linking the marketplace storefront as the order destination. Always your own surface, where it exists.

#7 — Hours

Including holiday hours, two weeks in advance. Christmas Eve. New Year's Day. Bank holidays. Local saints' days that matter in your market. The model penalizes venues that publish stale hours or that show "hours might differ" warnings — these warnings appear when you have not updated holiday hours in advance.

Operating ritual we install: every Monday, look at the next 14 days. Adjust any hours that need adjusting. 90 seconds of work, weekly. Compounds.

#8 — Posts (weekly cadence)

The single most under-utilized engagement surface. Yes, weekly. We have not seen a Google Business strategy work without it.

Post format we recommend, on a 4-week rotating template:

• Week 1: Menu highlight (a new dish, a seasonal special).
• Week 2: Offer (a specific, time-bound incentive — not "great food!").
• Week 3: Event (a wine dinner, a chef collaboration, a tasting).
• Week 4: Behind-the-scenes (the team, the supplier, the kitchen — humanize).

Each post takes 12 minutes to draft and publish. We log the post on a 12-week editorial calendar so the team never sits down to a blank canvas. The cumulative effect over six months in our portfolio: between 14% and 28% lift in profile views and a substantial increase in direct-website-click conversions.

#9 — Photos

Minimum 30 high-quality images, refreshed monthly with at least four new. Variety matters: dishes, drinks, room, team, exterior, golden-hour shots, daytime shots, busy-night atmosphere.

The model uses photo recency as a signal of venue vitality. Profiles whose newest photo is more than 90 days old are read as "potentially closed or inactive" and quietly downgraded.

If you are running a multi-location group, each location's profile needs its own unique photos. Shared brand assets across locations is read as a chain pattern, which can be a citation disadvantage in independent-restaurant queries.

#10 — Q&A — populate the questions yourself

A widely overlooked surface. Diners ask questions in this section, and Google publishes the answers. If you do not answer, anyone can — including competitors. We have seen profiles where the most-upvoted answer to "do you take reservations?" was an incorrect "no" posted by a random user.

The play: pre-seed the section with the 12 most common questions for your venue, answered by yourself. Do this once. Then monitor weekly for new questions and respond within 24 hours.

#11 — Attributes — be exhaustive

Wi-Fi, accessibility (wheelchair-accessible entrance, restroom, parking), payment methods (cards, contactless, NFC mobile), ambience (casual, romantic, cozy, upscale), good for (groups, kids, dates, business meetings), dietary (vegan options, vegetarian, gluten-free), serving (breakfast, brunch, lunch, dinner, late-night, drinks).

Every attribute is a potential filter the AI layer applies to compose answers. The model cannot infer attributes from your photos or your reviews with high confidence. It can only filter on what you explicitly declare. Declare everything that is true.

Why this is now critical

Google's AI Overview is reading every one of these eleven fields, every time it composes a restaurant answer. Profiles that look "complete" by 2022 standards look thin by 2026 standards. The diner asking AI "where can I take my parents tonight that has a quiet patio and takes reservations and offers vegetarian options?" gets an answer composed from the profiles that filled all eleven.

If you filled three of the eleven and your competitor filled all eleven, the citation goes to the competitor — even if your reviews are better, even if your food is better, even if you are physically closer to the diner.

The work pattern

Pick one settings batch per week. Touch the profile every Monday for 20 minutes. Photos refreshed monthly. Posts published every Thursday before dinner service. Q&A monitored weekly.

This is roughly two and a half hours of work per month, per location. The lift on profile views and direct conversions, in our portfolio, is between 18% and 41% within two quarters. It is the highest return on operator time we have ever measured in restaurant marketing — by a wide margin.

This is the kind of operational rigor that compounds, and the reason we run it as a managed service for every restaurant on the Kitxens platform. Whether you run it yourself or with a partner, run it. The diners are searching. The model is answering. Be the answer.

#GoogleBusinessProfile #LocalSEO #RestaurantMarketing #LocalSearch #GoogleMaps #AISearch #RestaurantDiscovery #SmallBusinessMarketing #HospitalityMarketing #Kitxens