Google Maps doesn't have a search bar anymore. It has an AI that's read every review.

Last week I opened Google Maps to find somewhere to take my wife for dinner, started typing, and caught myself doing the old thing: "italian surry hills". Keywords. The way we've all searched maps for fifteen years.

Except that's not really how Google wants you to search anymore. On 12 March 2026, Google announced Ask Maps, a Gemini-powered feature that lets you describe what you actually want in plain language instead of guessing the right keywords. The query Google itself used in the announcement is a good one: "My phone is dying, where can I charge it without waiting in a long coffee line?"

Read that again. There's no keyword in there to match. No business is going to have "phone charging, no coffee queue" in its name or category. The AI has to read across reviews, attributes, and descriptions to work out which places fit a situation a human described in a sentence. That's a different machine doing a different job.

For anyone running a local business, or anyone like us at Webcoda who advises them, this is the thing worth paying attention to. Not because it's flashy. Because it quietly changes what you're optimising for.

What Ask Maps actually does

Here's Google's own framing of it.

Google Maps

@googlemaps

Sometimes you need to find a place based on a specific, nuanced need. By asking exactly what you’re looking for, Google Maps will use Gemini to connect the dots across millions of reviews, then proactively show you insider tips from real people. For example:

791

12 Mar 2026

The key phrase is "connect the dots across millions of reviews". Maps has always had reviews. What's new is that Gemini reads them as evidence to answer a specific, messy human question, rather than treating them as a star average and a wall of text you scroll past.

So instead of "restaurant italian", you ask "somewhere quiet enough to actually talk, not too pricey, that does proper pasta". Instead of "mechanic near me", you ask "who can look at a brake noise today without me booking a week out". The query carries intent, context, and constraints all at once, and the AI tries to satisfy all of them.

Google Maps has more than 2 billion monthly active users, so this isn't a fringe experiment. It's a change to the front door of how a very large number of people find places. The rollout started in the US and India, on iOS and Android. Real users have been reasonably positive about it, which is worth noting because most of the loud commentary has come from SEO professionals rather than ordinary people.

Ivan Leo

@ivanleomk

@adelwu_ Tbh Gemini is really good at just finding spots haha too with ask maps. It’s worked quite well nowadays for me in sf

5

29 Apr 2026

I find that kind of low-key "yeah, it just works" feedback more telling than the breathless takes. It suggests the feature does the basic job well, which is exactly when a change like this becomes permanent rather than a gimmick Google quietly retires.

The shift that actually matters: from keyword to use case

If you take one idea from this article, take this one. The clearest articulation I've seen of what's changed came from a developer building local search tooling, and it's worth quoting directly.

Vikram | Building LocaTrack

@LocaTrack

Ask Maps lets users describe a situation instead of typing keywords. Quiet enough for a date. Dog friendly patio with shade. Schema and reviews now have to answer questions, not list features. The optimization unit is the use case, not the keyword. #LocalSEO #SEO #GEO

0

28 Apr 2026

"The optimisation unit is the use case, not the keyword."

That's the whole thing in nine words. For fifteen years, local SEO was a keyword game. You wanted your business to match the words people typed. You stuffed your category, your business name where you could get away with it, your description. The unit of optimisation was the search term.

Ask Maps breaks that model because people aren't typing search terms anymore. They're describing situations. "Quiet enough for a date." "Dog-friendly patio with shade." "Somewhere my toddler won't get glared at." None of those are keywords you can target. They're use cases the AI has to infer from whatever evidence it can find about your business.

So the question changes from "do my listings contain the right words?" to "does the evidence about my business answer the questions people are actually asking?" Those sound similar. They're not. One is about matching strings. The other is about whether your reviews, attributes, and profile actually describe what it's like to be a customer.

This matters because it moves the centre of gravity. Your business name and your chosen category were always under your control. Your reviews mostly aren't. If the AI is weighting review sentiment heavily (and the people watching this closely think it is), then the thing that decides whether you show up is partly out of your hands. That's uncomfortable, and it's also why the businesses that start now will have an advantage. Reviews accumulate slowly. You can't cram for this the night before.

How a business gets picked by the AI

Nobody outside Google knows the exact weighting, and I'm not going to pretend otherwise. But the mechanics are reasonably easy to reason about, because Gemini is working from the data Google already has on your business.

There are roughly four inputs the AI has to work with:

Your Google Business Profile fields. Categories, attributes, services, hours, description. This is the structured stuff. Attributes especially: "outdoor seating", "wheelchair accessible", "free wifi", "good for kids", "dog friendly". These map almost directly onto the constraints people put in conversational queries. If someone asks for a dog-friendly patio and you've ticked "dog friendly" and "outdoor seating", you've made the AI's job trivial. If you've left those fields blank, you've made yourself invisible to that query even if you've got the best dog-friendly patio in town.

Your reviews, as text. Not the star average. The actual words. This is where the use-case matching happens. A review that says "perfect quiet spot for a first date, we could actually hear each other" is gold for a whole category of queries you could never have targeted with keywords. A review that says "5 stars great" tells the AI nothing.

Your photos. Gemini is multimodal, so images aren't just decoration anymore. A photo of a shaded courtyard is evidence for "shaded patio" in a way a text field can't fully capture.

Freshness and relevance. Recent reviews, current photos, accurate hours. Stale data is a liability when the AI is trying to give a confident answer.

Here's the uncomfortable bit for anyone who's been coasting on a good star rating. Your five stars don't help much if none of your reviews mention the thing the AI is looking for. A restaurant with a 4.9 average and forty reviews that all say "amazing!" is weaker, for conversational search, than a 4.5-star place with reviews describing the gnocchi, the allergy handling, and the quiet corner table. The first one is praised. The second one is described. The AI can use description. It can't do much with praise.

(I went and checked Webcoda's own Google Business Profile while researching this, fully expecting to be smug about it. We had gaps. Attributes unticked, a couple of stale photos. If we're advising clients on this, that's a bit rich. It's on the list to fix this week, which is roughly the energy I'd suggest everyone bring to it.)

What this means for Australian businesses

The honest starting point: as of June 2026, Google hadn't confirmed when Ask Maps lands in Australia. It started in the US and India. Google rolls these things out progressively, so an Australian release sometime in the next year or so is a reasonable bet, but I'm not going to quote you a date Google hasn't given.

That uncertainty is actually fine, because the preparation is the same regardless of the date, and it pays off in normal Maps search too. There's no downside to a complete, well-reviewed profile.

The most concrete Australian angle I found was about tradies, which makes sense, because trades are where review sentiment and responsiveness already decide who gets the call.

Quick Suite Australia

@QuickAppAU

Google ‘Ask Maps’ AI is the new edge for Aussie tradies. 🛠️

Google ranks you on review sentiment now. Speed + quality = local dominance. 📈

quick reviews handles the heavy lifting.

Dominate local search: quickreviews.com.au ⚡

0

30 Mar 2026

The framing there ("speed plus quality equals local dominance") is marketing-speak, but the underlying point is sound for Australian service businesses. If you're a plumber, electrician, mechanic, or any trade where people search in a panic ("burst pipe, who can come today"), conversational search rewards exactly the things that already win you work: fast response, reviews that describe the actual job, and a profile that says what you do and where.

So here's the practical checklist I'd give any Australian SMB right now, Ask Maps or no Ask Maps:

• Fill out every Google Business Profile field. Categories, services, attributes, hours, the lot. Blank fields are missed queries.
• Tick every attribute that's genuinely true. Outdoor seating, wheelchair access, accepts EFTPOS, free parking, dog friendly. These line up directly with how people phrase constraints.
• Change how you ask for reviews. Stop asking for "a quick five stars". Ask customers to mention what they came in for and how it went. "If you've got a sec, it really helps if you mention what you had and whether the booking was easy." You want description, not just a rating.
• Respond to reviews. Your replies are text the AI can read too, and they signal an active, current business.
• Keep photos fresh. Current, specific, showing the actual space and what you offer.

The industries most exposed are the obvious ones: hospitality, retail, trades and services, healthcare, anything where people choose between options based on fit rather than just price. If your listing says "Italian restaurant" and nothing else, the AI has very little to work with. If your reviews say "the gnocchi's excellent, they sorted out a dairy allergy without fuss, and there's a quiet patio out the back", the AI can match you to a dozen different conversational queries you could never have targeted manually.

The counterargument worth taking seriously

I should be straight about the hype here, because a lot of the noise around Ask Maps is coming from people selling local SEO services, and a panicked market is good for that business.

The loudest claim doing the rounds is that Ask Maps "skips 90% of businesses". That number comes from an SEO consultant's promotional post. It's not Google's figure, it's not from a study, and as far as I can tell it's not from anywhere except a content marketing strategy. So I'd ignore the specific number entirely.

The honest version of that concern is more modest, and still real. When an AI gives you a short curated answer instead of a scrollable list of twenty pins, fewer businesses get surfaced per query. That's just structural. A list shows many; a recommendation shows few. So visibility probably does concentrate somewhat. But "somewhat" is not "90%", and pretending otherwise to sell a service is the kind of thing that's earned the SEO industry its reputation.

There's a second counterargument worth sitting with: maybe this is overblown because keyword search isn't actually going anywhere. Google didn't delete the search bar. Plenty of people will keep typing "cafe near me" because it's fast and it works. Ask Maps is an additional way in, not necessarily a replacement, at least not yet. If that's how it plays out, then "the keyword is dead" is too strong, and the right framing is "there's a new way to be found, and it rewards a richer profile". I think that's closer to the truth than the apocalyptic version. The advice doesn't change either way. A complete, well-described profile wins under both models.

The bigger pattern

Ask Maps doesn't exist in isolation. It's the same shift hitting every search surface at once. Google's AI Overviews changed regular search. ChatGPT Search and Perplexity changed how people research. Now Maps. The common thread is that every interface that used to hand you a list of results is being replaced by an AI that hands you an answer.

We've written about this from the web-search side before, and the principles transfer directly.

Related Article9 min read

AEO vs SEO: How AI Is Changing Google Search [2025 Australian Guide]

AEO is replacing SEO. AI search engines give direct answers, not links. Here's how Australian businesses must adapt.

Read full article

The underlying skill, making sure your content actually answers the questions people ask rather than just containing the right keywords, is the same skill whether the AI is summarising web pages or restaurant reviews. Structured, specific, current information wins. Vague keyword-stuffed information loses.

Related Article9 min read

Your Rankings Are Slipping. AI Changed Google's Rules.

You've done everything right. Keywords, backlinks, content. But your rankings keep dropping. Here's the AI-era signal your competitors are using that...

Read full article

If you've already done the work on structured data and answer-focused content for your website, you're ahead, because the mental model is identical. Ask Maps is just that same model applied to your physical business and its reviews instead of your web pages.

Where this is heading

Here's the direction, plainly: optimising for conversational and use-case queries is going to become a standard, explicitly-named line item in local SEO retainers across the Australian agency market, not buried under "Google Business Profile management".

I'm reasonably confident, because the incentive is clear. The moment Ask Maps reaches Australia (whenever that is), agencies will package "AI local search readiness" as a service, because that's what agencies do with every new Google feature. The interesting question isn't whether it becomes a line item. It's whether the work behind that line item is real, or just the same profile tidy-up rebadged with "AI" in front of it.

For most businesses, the practical answer is the boring one. Complete your profile. Tick your attributes. Ask for reviews that describe the experience, not just rate it. Keep your photos current. None of that is exciting. Neither was "get a website" in 2005, and that one aged reasonably well.

I don't know exactly when this lands in Australia, and I don't know precisely how much visibility concentrates when it does. Nobody outside Google does yet. But the preparation costs you almost nothing, it helps you under the search model that exists today, and it'll help you more under the one that's coming. That's about as close to a no-regret move as this industry ever offers.

---