AI Local SEO: What Actually Changes for a Local Business

The most useful finding in local search right now is a contradiction. AI Overviews appear on roughly two thirds of local business queries — and almost never on the ones where someone is about to walk through your door.

Whitespark ran 540 manual queries across three US cities and six industries and split them by intent. For local-intent searches like “personal injury lawyers in Phoenix”, a local pack appeared 93% of the time and an AI Overview only 15% of the time. For informational searches in the same industries, the ratio inverted almost exactly: an AI Overview 92% of the time, a local pack 6%.

So AI has not taken the “plumber near me” query. It has taken the research query that happens twenty minutes before it — how much this normally costs, whether you need a professional at all, what to ask before hiring one. That is where your visibility is genuinely at risk, and it is a completely different content problem from ranking in the map pack.

This guide covers both halves of what people mean by “AI local SEO”: being found by AI for local queries, and using AI to do local SEO work without breaking anything.

The intent split, and what it means for where you spend

Source: Whitespark, 540 manual queries, three US cities, six industries, May 2025. A deliberately narrow sample — treat the direction as reliable and the exact percentages as indicative.

The practical reading is that a local business now has two visibility problems that were previously one. The map pack problem is largely unchanged and still answered by profile quality, review volume and proximity. The new problem is that the informational content which used to feed your funnel is being summarised, and the summary often cites a directory, a national publisher or a forum rather than you.

Whitespark’s own citation analysis found the pattern starkly in one dataset: roughly 60% of AI citations in a plumber query set pointed to third-party publishers such as Reddit, Indeed, Yelp and Thumbtack, and about 40% to individual businesses. For a small local operator, that is the whole strategic issue in one number.

The concentration risk worth understanding

There is a second finding circulating that deserves attention with a caveat attached. Sterling Sky’s tracking, reported through several local-search publications, found that AI-generated local packs surfaced 5,943 unique businesses across a query set where the classic three-pack surfaced 18,330 — roughly a third of the coverage.

We have not seen the underlying methodology published in full, so treat the ratio as indicative rather than settled. But the mechanism behind it is easy to believe: a three-pack rotates through nearby businesses as the searcher moves, while a generated answer names whichever handful of businesses the sources happen to be most confident about. Fewer slots, and less rotation.

If that holds, the winners are businesses that appear consistently and accurately across many sources, and the losers are competent businesses with thin footprints outside their own website. That is not a new principle in local SEO — it is citation consistency with higher stakes.

What actually changed on Google Business Profile

Two changes in the past year matter more than the rest, and one of them removed a feature you may still be maintaining.

Q&A is being replaced by generated answers. Google announced in September 2025 that it was shutting down the Business Profile Q&A API, which was discontinued on 3 November 2025, with deprecation of the public-facing Q&A section beginning 3 December 2025 and rolling out gradually. In its place, Google has been rolling out an Ask Maps experience in which Gemini answers questions about a business by drawing on the profile, its reviews, its photos and its website.

The implication is direct and often missed. The answers customers get about your opening hours, parking, accessibility, languages spoken or whether you take walk-ins are now generated from your profile and reviews rather than typed by you. If your profile is thin or your website contradicts it, the generated answer will be wrong, and you no longer have a Q&A field in which to correct it. The fix is upstream: make the source material accurate and complete.

The profile is now an input to AI answers generally. Google’s own guidance for performing well in AI experiences tells site owners to think multimodal and to ensure their Merchant Center and Google Business Profile listings are updated. Your profile is no longer only a Maps asset. It is a structured, trusted, machine-readable description of your business that AI surfaces read. A neglected profile is now an AI visibility problem as well as a local ranking one — which is a genuinely new reason to do the work covered in our Google Business Profile optimisation guide.

Where AI genuinely helps the local SEO workflow

The second sense of “AI local SEO” is using the tools rather than being found by them. Here the honest assessment is that AI is a strong drafting and analysis assistant and a poor source of facts about your own business.

The pattern is consistent: AI is good where the input is your own real data and the output is language, and bad where the model has to supply the facts. That is the whole rule, and it holds across every tool in the category. What the tools cost and which are worth paying for is covered in AI SEO tools for small businesses.

The line you must not cross, and why Singapore is stricter than most

Generative tools have made fake reviews trivially cheap to produce. They have not made them legal, and Singapore is a poor place to test that.

Google’s own prohibited content policy is unambiguous. Content must be “based on a real experience” and must “accurately represent the location or product”. Reviews “that have been paid for, directly or in kind” are banned, as is posting “from multiple accounts by or at the request of one person”. Merchants specifically cannot “offer incentives — such as payment, discounts, free goods and/or services — in exchange for posting any review or revision or removal of a negative review”.

That last clause catches a practice many Singapore SMEs still run: the “leave us a five-star review and get 10% off” card at the counter. It is a policy violation regardless of whether the review is genuine.

Singapore adds a legal layer on top. Fake positive reviews can amount to an unfair practice under the Consumer Protection (Fair Trading) Act, which prohibits false or misleading claims. This is not theoretical: the CCCS has publicised findings against businesses for unfair practices centred on misleading web design features and the posting of fake customer reviews, and has secured a jail sentence against a managing director for failing to comply with court orders relating to CPFTA unfair practices. Separately, the Singapore Code of Advertising Practice administered by ASAS covers misleading advertising claims, and ASAS can require an advertisement to be amended or withdrawn.

The compliant version of the same goal is dull and effective: ask every customer, ask promptly, ask without conditions, and make it one tap. Our guide to getting Google reviews in Singapore covers the mechanics. Note also that if you are handling customer contact details to request reviews, that is personal data and the PDPA applies — the consent and purpose rules are covered in our PDPA and marketing tracking guide.

What a Singapore local business should actually do

Singapore has an unusual profile for this work. Google handles 92.46% of local search as of July 2026 per StatCounter, so Google’s surfaces dominate. Consumer generative-AI adoption is near the top globally at roughly 61% according to Stanford HAI’s 2026 AI Index, so assistants are genuinely in use. And the country is small enough that proximity — the dominant local pack factor almost everywhere — discriminates far less than it does in a large city. In a market you can cross in forty minutes, “near me” is a weaker filter, and reputation, specificity and content do more of the work.

A sensible sequence:

Get the profile genuinely complete. Correct primary category, every applicable secondary category, real hours including public holidays, services, attributes, current photos. This is now feeding generated answers, not just the map pack.

Make your website agree with your profile. Generated answers reconcile the two. Contradictions produce confidently wrong answers about your business, and there is no Q&A field left in which to correct them.

Build the informational content you have been neglecting. This is where AI Overviews are taking the impression, and where a page answering the real question — what this costs in Singapore, what the process involves, what to check before choosing — can earn a citation your competitors have not tried for.

Fix your presence on the third-party sites that rank in your category. Accurate, current, non-manipulated entries on the directories and comparison sites that genuinely appear for your terms. Given that a majority of local AI citations in the Whitespark dataset went to third-party publishers, this is no longer optional hygiene.

Ask ten local questions monthly, and log the answers. Put the questions your customers ask to Google’s AI surfaces and to one assistant, and record whether you are named, whether a competitor is, and whether anything said about you is wrong. Wrong is the finding that matters most, because it is fixable and nobody else is checking.

Keep the “near me” fundamentals running. They still decide the transactional query, which is still where the revenue is — see our guide to near-me searches in Singapore.

How to measure any of this

Business Profile insights still report calls, direction requests and website clicks, and those remain the cleanest local signals you have. In GA4, remember that traffic from Google’s AI Overviews and AI Mode is classified as Organic Search, not as AI — only chatbot referrals with a medium matching ai-assistant land in the AI Assistants channel, and a large share of assistant referrals arrive with no referrer at all and fall into Direct. The reports that matter are in our GA4 guide.

The single most reliable local measurement remains the least technical one: ask new customers how they found you, and write it down. In a market this size, fifty answers tell you more than any dashboard.

Frequently asked questions

Is AI going to replace the local pack?

There is no sign of it on transactional queries. Whitespark’s data shows a local pack on 93% of local-intent searches against an AI Overview on 15%. Where AI has taken over is informational and hybrid queries — 92% and 97% respectively. Expect the map pack to remain the mechanism for “find me one nearby”, and expect the research stage before it to be answered without a click.

Do I need to do anything different to my Google Business Profile because of AI?

Not different, but more thoroughly and more accurately. Google’s guidance for AI experiences explicitly tells businesses to keep Business Profile and Merchant Center listings updated, and the Q&A section is being replaced by AI-generated answers drawn from your profile, reviews, photos and website. That means gaps and contradictions now produce wrong answers rather than blank ones, and you have lost the field you used to correct them in.

Can I use AI to write my Google reviews or reply to them?

Replies, yes, with editing — that is your own voice responding to a real customer. Reviews, absolutely not. Google’s policy requires content based on a real experience and bans reviews posted at the request of one person, and in Singapore fake positive reviews can constitute an unfair practice under the CPFTA, an area where the CCCS has actively enforced. Offering a discount in exchange for a review is also a policy breach, even if the review is honest.

Does my business need to appear on Reddit or directory listicles now?

Accurate presence on the directories and comparison sites that genuinely rank in your category is worth having, since a majority of local AI citations in the Whitespark dataset went to third-party publishers rather than to businesses. Manufacturing forum threads about yourself is a different thing and is both detectable and, if the commercial relationship is undisclosed, a consumer-protection risk in Singapore.

Is proximity still the biggest local ranking factor in Singapore?

It is consistently ranked among the top factors in practitioner surveys of local search, but Singapore blunts it. In a market small enough to cross in under an hour, distance separates competitors far less than it does across a sprawling city, so review quality, category accuracy and content depth carry proportionally more weight here than the international guidance implies.

Should I pay for an AI local SEO package?

Not as a separate product. Google states there are no additional requirements or special optimisations for its AI features, and the work that matters — profile completeness, review generation, informational content, third-party accuracy — is local SEO as it has always been practised. Adding monitoring of what assistants say about you is a reasonable line item. Rebranding the whole engagement is not.

Where this leaves you

The comforting part of the local picture is that the query closest to the sale has barely moved. The uncomfortable part is that the research query feeding it is now often answered without anyone visiting a website — and when a source is named, it is more likely to be a directory or a forum than a local business.

So the work splits cleanly. Keep the profile, reviews and proximity fundamentals sharp, because they still decide the transactional query. Then build the informational content you have probably been skipping, because that is the surface being redistributed, and it is the one where a specific, local, honestly-sourced answer can beat a national publisher.

If you want that run properly, our AI SEO team in Singapore handles it inside an SEO engagement rather than as a separate package, and the results we have delivered for other local businesses are in our case studies. Related reading: how to show up in AI Overviews, how to rank on ChatGPT and the other assistants, and the Singapore AI SEO guide for the strategic overview.

Last updated 4 August 2026. Written by Adrian Tan and the SDM team. Sources: Whitespark, “The prevalence of AI Overviews in local search”, 540 queries, May 2025; Whitespark 2026 Local Search Ranking Factors survey (November 2025); Sterling Sky AI local pack coverage tracking, as reported in local search publications; Google Business Profile API change log and sunset dates documentation; Google Search Central, “AI features and your website” and “Top ways to ensure your content performs well in Google’s AI experiences on Search”; Google Maps prohibited and restricted content policy; Consumer Protection (Fair Trading) Act and CCCS enforcement reporting; Singapore Code of Advertising Practice, administered by ASAS; StatCounter Singapore search engine market share, July 2026; Stanford HAI 2026 AI Index.