Local AEO: How to Show Up When Customers Ask AI to Recommend a Business Near Them

Local AEO is the work of getting your business named when someone asks an AI assistant to recommend a company near them, for prompts like “best physio in Leicester” or “recommend a reliable electrician in LE2”. It sits exactly where answer engine optimisation meets local SEO: the same goal as national AEO (be the answer, not just a link), but the signals that decide it are local ones. Your Google Business Profile, your reviews, your listings in local and trade directories, consistent name-address-phone data, and the map that anchors you to a place.

Here is why it matters right now. In 2026, 45% of consumers say they use AI to find local businesses, up from just 6% a year earlier, which makes AI the third most-used source for local recommendations behind only Google and Facebook (BrightLocal, Local Consumer Review Survey 2026). This guide is the bridge between the answer engine optimisation pillar and everything in local SEO. It covers what local AEO is, how assistants actually choose a business, the uncomfortable odds, the six-step playbook in the order that works, and how to measure whether you are being recommended at all.

What is local AEO, and how is it different from national AEO?

Local AEO is answer engine optimisation for searches with a place attached. National AEO is about becoming the cited answer to a general question (“what is the best project management tool for a small studio”), and it runs mostly on authoritative content, clear entity data and being referenced across the web. Local AEO is about becoming the recommended business for a question with a location, spoken or implied (“near me”, “in Leicester”, “in LE1”).

The real difference is the fuel. A brilliant blog post will not get a plumber recommended in Wigston. What gets that plumber recommended is a complete Google Business Profile, a stack of recent reviews, consistent listings, and clear location signals the assistant can corroborate. National AEO leans on your ideas; local AEO leans on your record as a real business in a real place.

That is why we treat this as one job, not two. Local SEO and local AEO share the same foundations, and the businesses being surfaced in AI answers are, overwhelmingly, the ones that already did the local groundwork properly. If you want the wider strategy first, the pillar guide on AEO sets the scene; this piece is the local half of it.

How do AI assistants decide which local business to recommend?

They build a shortlist from a handful of local trust signals, then name the two or three businesses that every source agrees on. When you ask ChatGPT, Gemini, Perplexity or Google’s AI for a business near you, it is not scanning a live map on your behalf the way you would in Google Maps. It is assembling an answer from the sources it trusts for local facts.

In practice, those sources are:

• Google Business Profile: your categories, services, attributes, opening hours, photos, Q&A and the themes in your reviews.
• Reviews: the volume, recency, rating and actual wording, across Google and third-party platforms.
• Local and sector directories: trade registers, town directories and “best of” lists.
• Consistent entity data: your name, address and phone number matching everywhere they appear.
• Map data: Gemini in particular is grounded directly in Google Maps, which is why its business facts were around 100% accurate in SOCi’s 2026 testing, against roughly 68% on ChatGPT and Perplexity (reported by Search Engine Land).

The logic underneath is corroboration. The assistant is effectively cross-checking your story. If your profile says one address, a directory says another and your website shows a third, the model has no confident, agreed-upon fact to state, so it reaches for a competitor whose details line up everywhere. Getting recommended is less about shouting louder and more about being consistently, boringly correct.

The assistants also weight these sources differently. Gemini leans hardest on Google Maps, ChatGPT and Perplexity blend what they learned in training with live retrieval from directories and review sites, and Google’s AI sits on top of its own local index. You cannot optimise for one and ignore the rest, which is the good news in disguise: the way to cover all of them is to strengthen the shared foundations they draw from. The general mechanics of that are covered in how to get recommended by ChatGPT; here we are focused on the local layer specifically.

The uncomfortable maths: AI recommends barely 1% of local businesses

AI is far stingier with local recommendations than Google is. SOCi’s 2026 Local Visibility Index analysed performance data from nearly 350,000 locations across 2,751 multi-location brands and found ChatGPT recommended just 1.2% of them, Perplexity 7.4% and Gemini 11%, against 35.9% for Google’s traditional local 3-pack (Search Engine Land). Put plainly, being visible in AI is three to 30 times harder than ranking well in local search.

Worse, a strong Maps position no longer buys you a seat at the table. In the retail data, only 45% of the top 20 brands by traditional local visibility also appeared in the top 20 most frequently recommended by AI. The overlap is partial, which means you can be doing everything right for the map pack and still be invisible the moment a customer phrases the same question to an assistant.

We will be straight with you here, because it would be odd to sell this and pretend otherwise. When we ran the AI-visibility audit on our own site, thesecondfloor.io, we scored 0 out of 48. Zero. We are a Leicester marketing agency and the assistants did not name us for the services we sell. That result is exactly why we built this cluster and started fixing our own house first, from the same starting line most local businesses are standing on.

The stakes are not theoretical. With 45% of consumers now asking AI for local recommendations (BrightLocal), sitting in the invisible 99% is not a rounding error. It is enquiries quietly going to whichever business the machine does choose to name.

The local AEO playbook, in order

Do these six things, in this order. The first two are non-negotiable foundations; skip them and the rest is decoration. None of it is exotic. It is local SEO done properly and made machine-readable, which is precisely what these systems reward.

1. Get your Google Business Profile pristine and complete

Your Google Business Profile is the single most important input to local AI answers, so complete every field. It is the structured, verified record the assistants (and the map Gemini reads) lean on hardest. Claim and verify it, set the most specific primary category you can, fill in your services with prices, write a description a human would actually read, set accurate opening hours, add real photos, and keep it active week to week. It is free, and Google states plainly that there is no way to pay for a better local ranking (Google Business Profile Help).

We wrote the full method in our Google Business Profile playbook. For local AEO, the headline is completeness and accuracy, because a half-finished profile simply gives the machine less to trust and less to repeat.

2. Make your name, address and phone identical everywhere

Pick one exact format for your NAP and use it to the letter across every profile, directory and page. This is the quiet lever that decides whether an assistant can state a confident fact about you. Remember that SOCi found business data only about 68% accurate on ChatGPT and Perplexity (Search Engine Land); inconsistency is a big part of why.

Decide now whether you are “Unit 4, 12 High Street” or “Suite 4, 12 High St”, whether the number is written 0116 or +44 116, and whether you spell out the full postcode. Then match that exact version on Google, Bing, Apple, your website footer and every listing you hold. A local geographic number signals locality better than an 0800 line, so use one as your primary. When your details agree everywhere, the model has something solid to say; when they clash, it moves on to a business it can describe without hedging.

3. Earn steady, recent local reviews

Reviews are how the algorithm, the AI, and the human reading its answer all sanity-check that you are any good. The locations AI recommends are not scraping the barrel: in SOCi’s data they averaged roughly 4.1 to 4.3 stars (Search Engine Land). And people do not take the assistant’s word blindly. Fully 88% of AI users verify a recommendation, most commonly by checking whether the reviews look legitimate (BrightLocal).

So a steady trickle of genuine, recent reviews does double duty: it lifts your local prominence and it survives the fact-check that follows the recommendation. Ask for them with a one-tap link at the moment a customer is happiest, reply to every one, and never fake or incentivise them. We cover the tactics and the UK legal position (fake and paid-for reviews are now illegal) in our reviews guide.

4. Get into the local and sector directories AI reads

Assistants corroborate you against directories and “best of” lists, so you need to be in the ones that carry weight. Language models lean on structured, third-party sources to decide who is real and who is worth naming, and a citation on a trusted directory is one of the clearest of those signals.

Quality beats quantity by a distance. A dozen accurate listings on well-known UK and sector-specific directories do more than a hundred entries on sites no human ever visits, and a scattering of inconsistent old listings actively works against you. Worth knowing: citations tend to count for more in AI answers than they do for the classic map pack, precisely because language models reach for structured third-party sources to decide who is worth naming. Prioritise the major players, then add two or three respected directories for your specific trade and your town. We keep a working list of the platforms that genuinely move AI recommendations in our guide to the UK directories that influence AI.

5. Publish location-specific, answer-shaped content

Give the assistants clear, local, question-shaped pages they can quote. Your website is the one part of this you own outright, and it is where you answer the real questions a local customer asks: what you do, which towns you cover, what it costs, and why you are a safe choice.

Write genuine location pages, not thin templates with the town name swapped in, and structure them so a specific passage answers a specific question. That passage-level clarity is exactly what gets pulled into an AI answer, because the model wants a tidy, quotable statement it can attribute. We go deep on the page craft in location pages that outrank competitors, and on the passage-level technique in how to rank in Google AI Overviews.

6. Add LocalBusiness schema so the machine can read you

Mark up your site with LocalBusiness structured data so search engines and assistants read your details without guessing. Schema translates your name, address, phone, opening hours, service area and geo-coordinates into a format machines parse directly, rather than having to infer them from your page copy.

It is not a magic ranking lever, and it will not rescue a thin profile. What it does is remove ambiguity, and ambiguity is the enemy of being confidently recommended. Pair clean schema with the consistent NAP from step two, and you have handed the systems a corroborated record straight from the source you control.

What good local AEO looks like in practice

Picture an independent physiotherapy clinic in Leicester that ranks on page one of Maps but never gets named when someone asks ChatGPT for “the best physio in Leicester”. This is an illustrative, composite example, not a specific client, but it is the shape of the job.

The work is entirely the six steps above. The profile is reclaimed and completed, with the primary category set to the specific service rather than a generic “physiotherapist”, and treatments listed with prices. The clinic’s name, address and phone are reconciled across Google, Bing, Apple, two health directories and the website footer, so every source tells the same story. A one-tap review request goes out after every discharge, listings are added to the right UK and physiotherapy-specific directories, three genuine neighbourhood pages answer real local questions, and LocalBusiness schema ships on the site.

A couple of months on, the clinic starts appearing in AI answers for its core services across the city, the details the assistants give are correct, and (the only figure that pays the bills) enquiries that mention “found you through ChatGPT” or Google’s AI begin turning up in the diary. The lesson is not that any single tactic is magic. It is that a complete, consistent, corroborated local presence beats a neglected one, reliably, over a quarter.

How do you measure your local AI visibility?

Build a category-plus-location prompt panel and run it every month. You cannot manage what you do not measure, and a single lucky check tells you nothing. A simple, repeatable method:

1. Write a fixed list of prompts a real customer would use: “best [your service] in [your town]”, “recommend a [your trade] near [landmark or postcode]”, “who should I call for [problem] in [town]”, plus a few “near me” variants.
2. Run them across the assistants that matter (ChatGPT, Google’s AI, Gemini and Perplexity), ideally from a location inside your service area, because local answers shift depending on where the query is asked.
3. Log three things each month: whether you are named at all, where you sit in the list, and whether the details the assistant gives about you are correct.

Watch the trend across months rather than reacting to one snapshot, and expect movement in two to three months after you fix the fundamentals, not two to three days. We wrote a full do-it-yourself version in how to check whether AI recommends your brand.

Pair that panel with the numbers you already have. Your Google Business Profile insights show calls, direction requests and website clicks, and simple call tracking tells you which enquiries turn into work. Together they let you connect a rising AI presence to something that matters rather than admiring a mention in isolation.

Keep one eye on the outcome that actually pays, too. Being named is the means; more of the right calls, forms and bookings is the point. A business that gets mentioned by three assistants but converts nobody has a different problem from one that gets mentioned once and books every enquiry.

Start here

If you do nothing else this month, complete your Google Business Profile and make your NAP identical everywhere. Those two fix the foundation the assistants trust most, and they are entirely within your control. After that, set up a weekly review request, claim your core directory listings, and add LocalBusiness schema to your site. That sequence alone moves most local businesses from invisible to occasionally-named within a quarter, because so few competitors have bothered to do it properly.

If you would rather not work through it alone, that is the work we do, and we started by running the audit on ourselves rather than preaching from a mountaintop. Book a free growth review and we will check your local AI visibility, your profile, your reviews and your listings, then hand you a prioritised list of what to fix first. Our engagements are scoped monthly, with weekly updates and no lock-in, so you can watch the needle move before committing to anything long-term. No jargon and no inflated promises, just a straight read on how to get recommended when your next customer asks an assistant who to call.