Ask Maps: Gemini is inside Google Maps now. What it means for local businesses

Ask Maps is Gemini built directly into Google Maps: you ask a full question, and Maps writes you an answer with recommended businesses, review snippets, and citations instead of showing a pin list. It launched March 12, 2026 in the US and India (Google, 2026), and for owners the implication fits in one sentence: your profile is no longer a listing, it is source material.

That distinction decides who wins the next few years of Maps. The old pin list read three things: name, category, distance. Ask Maps reads everything, and it quotes what it finds. This article covers how the surface works, what it reads, and the specific work that makes your business quotable.

What is Ask Maps, and what does a query look like?

It is conversational search on the highest-intent local surface there is. Google's own launch examples show the shape: "my phone is dying, where can I charge it without waiting in line" and "is there a public tennis court with lights I can play at tonight" (Google). Multi-constraint, situational, the way people actually talk. Maps parses the intent, filters by real data like hours and attributes, and answers in prose with businesses attached, then takes follow-up questions in the same thread (CNBC, 2026).

Availability note for Canadian readers: the rollout started with the US and India, mobile first, desktop coming (Search Engine Land, 2026). Google has not dated Canada, but AI Overviews and AI Mode both crossed the border within months. Treat this as prep time, not a pass.

Where do Ask Maps answers come from?

From the data Maps already holds, read deeper than any surface has read it before. Google says Ask Maps draws on over 300 million places and the contributions of more than 500 million people (Google). In practice, testing shows answers assembled from business profiles, review text, photos, Q&A, websites, and live signals like hours and busyness, stitched together with citations (GSQI, 2026).

That gap is the whole opportunity. The fields the AI leans on hardest, attributes, services, review detail, Q&A, are exactly the ones most owners never filled, because the old pin list never displayed them. Nobody bothered decorating rooms no customer could see. The rooms are visible now.

Why does review text now matter as much as stars?

Because the text is what gets quoted. When someone asks for "a patio quiet enough for a meeting," Ask Maps cannot learn quiet from a star count. It learns it from a reviewer who wrote "the back patio stays quiet even on Saturdays." Star ratings still gatekeep trust; the sentences underneath them are now retrieval data, surfaced directly in answers as review snippets and sentiment summaries (GSQI).

You cannot write your customers' reviews, but you can shape what they mention. Ask for the review right after the specific job, and ask the question that prompts detail: "would you mind mentioning what we did for you?" A homeowner who writes "replaced our furnace same-day in January" just taught the machine three answerable facts. The review system guide covers the mechanics, and your responses count too: replying with specifics ("glad the patio worked for your meeting") adds more indexable text to the pile.

What does a deep profile look like?

Deep means every field an answer could be built from is filled, honestly and specifically. The checklist, in the order the machine seems to care:

• Attributes, all of them. Outdoor seating, wheelchair access, Wi-Fi, parking, women-owned, dog-friendly, payment types. These are first-pass filters for multi-constraint questions. An unset attribute reads as a no.
• Services and menu items, named. "Furnace repair," "lash lift," "gluten-free crust." Ask Maps matches questions to listed services; a generic category cannot answer a specific ask.
• Photos of the things people ask about. The patio, the parking, the accessible entrance, the finished jobs. Photos confirm claims and get surfaced in answers.
• Q&A, answered by you. The questions sitting unanswered on your profile are literal question-answer training pairs for this surface. Answer them the way you would to a customer at the counter.
• Hours, holiday hours, and real-time accuracy. "Open past 9" filters on your stated hours. Wrong hours now exclude you from answers, silently.

Run the 15-minute profile audit with this lens and most owners find half the checklist empty. That is not a failure; it is headroom your competitors have not noticed yet.

How does Ask Maps fit the bigger AI arc?

It is the third landing of the same shift, and the most direct one yet. First, AI answers replaced the pin list in Search: AI local packs now feature one or two businesses inside written recommendations. Then assistants like ChatGPT became a discovery channel of their own, with 45% of consumers using AI tools to find local businesses in 2026, up from 6% (BrightLocal). Now the conversation has moved inside Maps itself, the app people open when they are minutes from spending money.

The strategic read: every one of these surfaces rewards the same underlying work. Whitespark's 2026 ranking factors report makes the point explicitly, the inputs that win Google, Maps, and the AI assistants have effectively merged (Whitespark, 2026). You are not optimizing for Ask Maps. You are building one deep, consistent, quotable presence that every surface, including the ones you can test today, reads from.

What should you do this month?

Four moves, none requiring a budget:

1. Fill the attribute and services gaps. Open your profile, go field by field, and set everything true. One sitting is fine here; adding data is not the risky edit pattern, changing core identity fields is.
2. Answer your Q&A. Every unanswered question on the profile, answered plainly. Then seed the two or three questions customers ask you weekly, and answer those too.
3. Point your review asks at specifics. Switch the ask from "leave us a review" to "mention what we did for you." Reply to new reviews with details that add context.
4. Photograph the askable things. The patio, the parking, the ramp, the work. Ten honest photos beat one hundred logo graphics on this surface.

Then verify. If Ask Maps is live where you are, run your buyer questions and see who gets quoted. If it is not yet, run the same questions in Gemini and AI Mode, which read the same data. There is no dedicated Ask Maps report; impressions fold into your normal profile analytics (GSQI), so your eyes are the tracking tool for now.

Frequently asked questions

Maps stopped being a list. Make your profile worth reading.

For twenty years, winning Maps meant winning a glance: closest pin, best stars, nicest thumbnail. Ask Maps replaced the glance with a read. Gemini goes through your attributes, your services, your customers' sentences, and your photos, and then it either quotes you or it quotes someone else. The businesses that treated their profile as a formality are about to find out what the formality was for.

The work is not mysterious and it is not expensive. It is an afternoon of filling fields, a review ask that prompts specifics, and photos of the things people ask about. If you would rather have it done properly and connected to a site the machines can also quote, that is an RMCM local SEO engagement, and the free audit shows you in 30 seconds how readable your business is right now.