Ask Maps now picks stores for shoppers. Test whether it picks yours

What is Ask Maps?

Ask Maps is a chat built into Google Maps. You type or say a whole question, and Google’s Gemini models answer with places they recommend ¹.

The answers draw on more than 300 million places, including reviews from more than 500 million contributors, and they’re personalized by places the person has searched for or saved in Maps ¹. From an answer, people can book a table, save a place, share it or get directions ¹.

To try it, open the Google Maps app on Android or iPhone and tap Ask Maps under the search bar. On a computer, open Maps, click Menu, then Ask Maps ².

For a store, that means Maps now does more of the choosing for the shopper.

What changed in August 2026

The dates and countries come from Google’s own posts ^1,3. Everywhere Ask Maps is available, the August update added live transit delays, conversation memory and Personal Intelligence, which lets people connect Gmail so suggestions account for their reservations. It’s off by default ³.

In the US only, “with more countries to come over time,” Ask Maps can now [3]:

• Order food, with Square and Toast at launch and Uber Eats coming soon ^3,4.
• Find hotels with live prices, and local events.
• Take edits and tips by chat. Someone can upload a photo of a storefront sign, and Maps will read the new hours from it and ask them to confirm before submitting.

Picture last year’s Thanksgiving sign still taped to your door. A shopper photographs it, and Maps turns it into a suggested edit to your hours. Google says its systems review suggested edits for accuracy ³. But once you’re notified of one, you have 4 days to accept or reject it, or Google may publish it if other public information, such as your website, supports it ⁵. Your sign, your profile and your website need to agree.

How does Ask Maps choose which stores to name?

Google hasn’t said. Its posts explain what Ask Maps can do, not how it picks or orders places. Search Engine Journal put it plainly: until Google shares more, claims about how Ask Maps ranks places “should be seen as educated guesses” ⁶.

What Google has published is how local results on Search and Maps work in general. They rest mainly on three things [7]:

• Relevance: how well a Business Profile matches what someone searched for.
• Distance: how far the business is from the person searching.
• Prominence: how well known a business is, based partly on links to it and on its reviews. More reviews and positive ratings can help.

The same page says complete, accurate profiles are more likely to show up, and nobody can pay for a better local ranking ⁷. Google’s attributes page adds that certain attributes might put you in searches for places with those attributes ⁸.

The right column is inference. It’s reasonable because Ask Maps answers from Maps’ own places and reviews ¹, and local search experts quoted by Search Engine Journal say it also draws on business websites and third-party sources ⁶. Treat it as a working theory to test.

The Business Profile fields behind each question

Every question needs a fact to answer it. Here’s where those facts live in your Google Business Profile, the free listing that puts your store on Search and Maps.

A note for chains: Google says all locations of a business must share one primary category, and each can add up to 9 more ¹⁰. So what sets store 14 apart comes from its extra categories, attributes, photos and description. Retail stores in eligible countries can also show in-store products, so shoppers can see what’s in stock ⁷.

Test it yourself: 10 questions for each store

You need a phone, one store’s facts and about 15 minutes.

Before you start:

1. Write down the store’s real facts: hours, Thanksgiving hours, the three things people come in for, parking, the entrance.
2. Name the neighborhood or a landmark in every question. Distance is part of local ranking ⁷, so don’t let your own location do the work.
3. Have someone who doesn’t work for you run the same questions on their phone. Answers are personalized ¹, and your phone has looked up your stores a hundred times.
4. Screenshot every answer and note the time. “Open now” depends on the clock.

The 10 questions. Swap in your own products, streets and landmarks.

1. “Where can I buy [product] near [neighborhood] that’s open right now?”
2. “Which [type of store] near [landmark] is open after 8 tonight?”
3. “Is [your brand] on [street] open on Thanksgiving?”
4. “I’m driving from [place] to [place]. Where can I pick up [product] on the way?”
5. “Which [type of store] near [neighborhood] has easy parking?”
6. “Is there a [type of store] near [neighborhood] with a wheelchair-accessible entrance?”
7. “Where can I get [a service you offer, like a key cut] near [neighborhood] today?”
8. “What do people say about [your brand] in [neighborhood]?”
9. “Which [your brand] store near [landmark] has [a department or product]?”
10. “Where’s a good place near [neighborhood] for [a reason people visit, like a last-minute gift]?”

Questions 1, 2, 4 and 10 test whether you’re named at all. The rest test single facts and how Maps describes you.

What to look for in each answer:

• Named, missing or swapped. Is your store there, and is it the right one?
• Wrong facts. Hours, address, what you sell, parking.
• Who’s named instead. Open the rival’s profile. Search Engine Journal suggests noting what the named places share, such as accurate hours, recent reviews, complete attributes and a website that explains what they offer ⁶.
• Words you didn’t write. If the answer says something your profile doesn’t, check the store’s page on your website first.

If an answer is wrong, tap the thumbs-down and describe the problem, as Google’s help page suggests ². The same questions work in ChatGPT and Google’s AI Overviews, which is the wider job of AI answers.

Store:                  Tested by:
Date and time:          Phone: mine / someone else's

For each of the 10 questions:
Q#  Named? yes / no / another of our stores
    Wrong facts:
    Named instead:
    What their profile has that ours doesn't:
    Fix, and who owns it:

Re-test on (two weeks out):

What to fix when a store is missing or described wrong

Match what you saw to the fix.

• Missing for what it sells. Check its extra categories and add in-store products if you can ⁷. Say in the description what people come in for. Google wants it about the business rather than promotions, with no links ¹⁰.
• Wrong hours. Fix the regular hours, then add special hours for every holiday. Google suggests confirming holiday hours even when they match your usual ones ⁹. Blue text in the profile editor marks info Google changed ⁵.
• Missing a detail the store has. Set parking, a step-free entrance or Wi-Fi as attributes. Google’s review of the edit usually takes about 10 minutes, sometimes up to 30 days ⁸.
• Described wrong. Rewrite this store’s description in its own words, and make its website page match. Search Engine Roundtable saw Google write an AI description of a business from old website text ¹¹.
• A rival wins on reviews. Reply to every review. Google says replies show you value feedback, and more reviews and positive ratings can help ⁷. Ask every customer, not only the happy ones. Here’s how reviews work at every store.
• The wrong one of your stores is named. Check both stores’ addresses, pins and descriptions. Our guess: two stores with the same description give Maps little to tell them apart.

Re-run the same questions in two weeks, on the same phones. One test is a snapshot, and answers move with the time of day and the person asking.

Where chains slip

At a September 2026 webinar featuring Google’s Caroline Dissaux, the speakers described searches getting longer and more conversational, with AI turning a broad local search into a specific request: a particular service, the right amenities, a convenient location ¹². Each of those lives in one store’s profile.

Four gaps are common, and all four are easy to check:

• One brand description pasted into every location.
• Holiday hours set from head office the week of.
• Reviews answered at the flagship and nowhere else.
• Attributes filled in at opening and never touched again.

A strong brand doesn’t fill in a blank attribute at store 14. Each store has to answer for itself, which is the everyday work of local search and listings.

Ordering from Maps: what Square and Toast restaurants should check

If you run restaurants or cafes in the US, Ask Maps can now take orders. A guest asks for a dish, Ask Maps finds open restaurants that serve it, even along their route, adds it to a cart, and the guest reviews and completes the order ³.

Square says its US food and beverage sellers who activate a Google Business Profile can show up in Ask Maps, with menus, hours and location details syncing from Square Dashboard and no added setup, contracts or fees. Orders flow into their Square point of sale and kitchen display ¹³. On Toast? Ask Toast what applies to your account.

• The Business Profile is verified and active at every location.
• Hours in your point of sale match the profile, holidays included.
• Menu items use the words guests say out loud. “Pad kee mao” beats “Noodle special 4.”

Google’s August post describes ordering from restaurants only, so a shoe store has nothing to switch on yet. Google says it is building this with partners on the Universal Commerce Protocol for Food ³, an open standard for agentic commerce, in Square’s words ¹³. For what AI agents that shop mean for retail, read agentic commerce.

A weekly Ask Maps check for district managers

Run all 10 questions once per store as a baseline. Then rotate three questions per store each week, so every store gets the full set about once a month.

• Open each store’s profile. Accept or reject suggested edits within 4 days of the notice, and look for blue text, which marks info Google changed ⁵.
• Check hours for the next 14 days. This season: Thanksgiving (Thu Nov 26), Black Friday (Fri Nov 27), Small Business Saturday (Sat Nov 28), Christmas Eve (Thu Dec 24), Christmas (Fri Dec 25), New Year’s Eve (Thu Dec 31) and New Year’s Day (Fri Jan 1).
• Ask three of the 10 questions per store and screenshot each answer.
• Reply to every review from the past week.
• Add one recent photo per store.
• Check one attribute against the real store, like the parking lot or the entrance.
• Log it: store, question, named or not, the fix, who owns it, the date.

Twenty stores times seven boxes is a lot of tapping. It’s the weekly work Kai, Ocally’s AI growth agent, does for every store: he reads each Google profile, compares it with your facts and drafts the fixes for your yes.

Whoever does it, do it every week. Ask Maps answers with whatever it can find about each store, and the profile is the part you control.