Google has expanded Ask Maps, the Gemini powered conversational layer inside Google Maps, to Australia, Brazil, Canada, Indonesia, Japan and Mexico, along with what it calls "over 150 countries and territories in English". Google published the announcement on 6 August 2026, and it takes Ask Maps well past the United States and India, the only two markets where the feature had been available before.
The rest of that announcement is where most summaries go wrong. Google also introduced food ordering, hotel discovery, event discovery and conversational contributions inside Ask Maps, and every one of those is rolling out in the United States first, with more countries to come over time. What actually crossed borders in August is the conversational answer layer itself, together with a live transit widget, personalised responses and conversation history.
What Google said on 6 August 2026
In its Maps blog post, Google set out two separate rollouts in the same piece of writing. The first is geographic. Ask Maps is, in Google's wording, "rolling out in Australia, Brazil, Canada, Indonesia, Japan, and Mexico, along with over 150 countries and territories in English". The second is functional, and it is held to one market at launch.
On the geographic side, Google said the live transit widget, personalised responses and conversation history are rolling out everywhere Ask Maps is available. The transit widget puts up to the minute delay information inside the assistant itself. Personalised responses come from securely connecting Ask Maps to Gmail, with Calendar named as a later addition, so the assistant can reason about a flight or a dinner reservation a person already has. Conversation history lets someone come back to an earlier thread and ask what the assistant suggested for a past trip, as long as history settings are switched on.
The agentic features are the ones Google led with, and they are the ones that stop at the United States border for now. Food ordering arrives, in Google's words, "with partners like Square and Toast, with Uber Eats coming soon". Hotel discovery, event discovery and conversational contributions, which let people add information back into Maps through the same chat interface, are also United States first.
The split that decides whether any of this touches your listing yet
Getting the split right matters more than the headline. A restaurant owner in Bangkok reading a summary that says Ask Maps now takes food orders in 150 countries will go looking for a feature that is not there. Google's own post separates the two rollouts clearly, and the table below keeps them apart. Every row comes from Google's 6 August post and nothing else.
| Feature | Where Google said it is rolling out |
|---|---|
| Ask Maps conversational layer | Australia, Brazil, Canada, Indonesia, Japan and Mexico, plus over 150 countries and territories in English. Previously United States and India only |
| Live transit widget, personalised responses, conversation history | Everywhere Ask Maps is available |
| Food ordering, with Square and Toast, Uber Eats coming soon | United States first, with more countries to come over time |
| Hotel discovery and event discovery | United States first, with more countries to come over time |
| Conversational contributions | United States first, with more countries to come over time |
How the expansion reached Southeast Asia inside a week
Indonesia is the one Southeast Asian country Google names directly in the 6 August post. Regional press filled in part of the rest. Malaysian outlets reported the Malaysia rollout around 7 to 10 August 2026, and Singapore coverage followed on 12 August 2026. That puts the conversational layer in three Southeast Asian markets inside roughly a week of the announcement, at least according to reporting on the ground rather than a Google country list.
The distinction is worth holding onto. Google named six countries and then used a bulk phrase for everything else. Anything beyond those six comes from regional reporting or from people opening the app and finding the feature, which is a weaker form of evidence than a published list, though it is the only form available for most markets right now.
Thailand is not named, and Google's post does not resolve it
Thailand does not appear anywhere in Google's 6 August post. The phrase "over 150 countries and territories in English" does not tell a Thai business two things it would want to know: whether Ask Maps is switched on for users in Thailand, and whether it answers in Thai at all. The wording points at English language availability, and Google did not publish a market by market breakdown to check against.
So the honest position is that the source does not say. The practical answer is to check rather than to assume. Open Google Maps on a mobile device with the app up to date, and look for the Ask Maps entry point in the app. If it is there, it is live for that account in that market. If it is not, no published Google document currently gives a date for Thailand. Treat any article that states a Thai launch date as unsourced until Google publishes one.
What a conversational local layer asks of a business profile
This is the part that has practical weight for anyone managing listings, and it holds whether or not Ask Maps is live in a given market yet. A conversational interface takes a question in ordinary language and has to resolve it against structured information about places. That is a different job from matching a query string to a business name, and it is worth reasoning about carefully, because Google did not describe how Ask Maps selects or ranks the places it mentions.
Consider the kind of question people put to a conversational assistant. "A quiet restaurant that is open past 10pm and has parking." "Somewhere near Asok that can seat eight people." "A clinic open on Sunday morning that speaks English." Each of those bundles several conditions into one sentence. To answer any of them, a system has to read opening hours, amenity and attribute fields, service details, category, and whatever text sits in reviews. A business whose profile is thin on those fields is harder to resolve against a compound question, no matter how strong its brand is.
Opening hours, including the awkward ones
Hours are the single field most likely to be wrong and most likely to be asked about. Regular weekly hours are usually filled in. Public holiday hours, Songkran closures, kitchen last orders that differ from closing time, and split hours for a lunch break are the ones that go stale. A question phrased as "open past 10pm" or "open now" can only be answered from that data. Set special hours ahead of Thai public holidays instead of after them.
Attributes and amenities
Attribute fields cover things like accessibility, parking, outdoor seating, wifi, payment methods, and whether a place is good for groups. These read like checkbox housekeeping, and for keyword matching they largely were. For a question that names a condition out loud, they are the only place the condition can be answered from. Go through the full attribute list for your category rather than the handful that were pre filled.
Services, menus and what you actually sell
Service and menu detail turns a category into specifics. A clinic that lists its procedures, a workshop that lists what it repairs, a restaurant with an actual menu attached, all give a conversational system something to match against a request for a named dish or a named treatment. A profile that says only "restaurant" answers far fewer questions than one that says what is on the menu and at what price.
Category selection
The primary category is the strongest single classification signal on a listing, and secondary categories cover the rest of what a business does. Businesses that changed what they sell over the past few years often never updated it. Check the primary category still describes the main line of business, and that secondary categories cover the real services rather than aspirational ones.
Photo coverage
Photos carry information that no text field holds: what the room looks like, whether it is quiet or loud, whether there is space for a pram, what the food actually arrives looking like. Coverage matters more than volume. Exterior, interior, seating, menu items, parking and the entrance are worth having, kept current rather than uploaded once at launch.
Review text and the words customers use
Review text is the only part of a listing written in the customer's own language. If people describe a place as quiet, or mention that they parked easily, or say the staff spoke English, that language exists in the profile without the business having written it. Asking customers for reviews is ordinary practice. Asking them to lie is not, and buying reviews is against Google's policies. The useful version of this is to make sure the experience people describe is the experience you want the profile to convey, and to reply to reviews in the same language the customer used.
What this means for Thai marketers
This section is analysis rather than anything Google stated. Bangkok, Chiang Mai and Phuket businesses run heavy local discovery volume, and much of it already runs through Maps rather than through a browser. If a conversational layer sits on top of that data, the profile stops being a listing and starts being the answer source. The work does not change much in kind, but the cost of an incomplete profile goes up, because a missing attribute is no longer a missed filter, it is a question the assistant cannot answer about your business.
Two practical notes for Thai listings specifically. First, business names, addresses and category names exist in both Thai and English on many profiles, and the quality of the two versions is often uneven. Second, Google's post frames the wide rollout as English language, so a Thai language question may behave differently from an English one, and Google published nothing about Thai language handling. Both of those are reasons to keep the English fields on a Thai profile as complete as the Thai ones. Solid local SEO work already covers most of this, and the same profile data feeds the wider set of AI answer surfaces that quote business information back to users.
What Google did not say
Being clear about the gaps is part of reading an announcement like this one. Google did not publish the full list of the 150 plus countries and territories. It did not name Thailand. It did not say whether Ask Maps responds in Thai or only in English in the wider markets. It gave no timeline for food ordering, hotel discovery or event discovery outside the United States beyond "more countries to come over time". It published no usage figures for Ask Maps, and it described nothing about how places are selected or ordered inside a conversational answer. Anyone telling you how to rank in Ask Maps is working from inference, not from this post.
Questions marketers are asking about Ask Maps
Is Ask Maps live in Thailand?
Google's 6 August post does not say. Thailand is not in the list of named countries, and the phrase "over 150 countries and territories in English" does not identify which markets are included. The way to find out is to open Google Maps on an up to date mobile app and look for the Ask Maps entry point.
Can customers order food from my restaurant through Ask Maps?
Not outside the United States, based on what Google published. Food ordering is rolling out in the United States first, with partners like Square and Toast and Uber Eats coming soon, and Google said only that more countries will follow over time with no date attached.
Do I have to do anything right now?
Nothing new is required, and nothing in the announcement creates a deadline. The work it points at is the work that was already sitting undone on most profiles: accurate hours including holidays, complete attributes, real service and menu detail, correct categories and current photos.
Does this replace local SEO?
No, and Google did not suggest it does. A conversational interface still reads the same underlying business data, so the maintenance job stays the same. What changes is the shape of the question being asked of that data, which rewards completeness across many fields rather than strength in a few.
Will my listing rank differently because of this?
Google published nothing about ranking in this post, so there is no sourced answer. Any claim about how Ask Maps orders or chooses the businesses it names is inference. What can be said is that a compound question needs several fields answered at once, and a profile missing those fields gives a conversational system less to work with.
If you want a second pair of eyes on how complete your business profile actually is against the kind of questions customers ask out loud, that is a short piece of work with a clear output, and it is worth doing before the agentic parts of Ask Maps reach this region.
