Quick answer: local AI shortlists are built from the same underlying record as the map pack — your Google Business Profile — but weighted differently. Proximity matters less, and consistency, review substance and third-party corroboration matter more. A business that dominates the map pack can still be missing from the assistant's answer.
What changes when the question goes to an assistant
A map pack ranks by relevance, distance and prominence, and shows three results with a map. An assistant asked "who should I use for X in Leeds" returns prose naming two to five businesses, often with a sentence of reasoning about each. That reasoning is the tell: the model is not just ranking, it is describing — and it can only describe what it has read about you somewhere.
Practically, this means a business with a thin but well-optimised profile can win the map pack and lose the assistant, because there is nothing to say about it beyond the category and a star rating.
The signals that decide a UK local shortlist
- NAP consistency across British sources. Name, address and phone identical on your site, Google Business Profile, Bing Places, Companies House, Yell, your trade body and any sector directory. Postcode formatting included — engines match on the exact string more often than you would like.
- Service-area clarity. If you cover Greater Manchester rather than Manchester city centre, say so explicitly in prose on the site, not only in the profile's service-area setting. The model reads the sentence.
- Reviews with substance. A review that names the job, the area and the outcome gives the model something to quote. A star rating with no text gives it nothing. Ask customers to mention what you did and where.
- Local corroboration. Regional press, local chamber of commerce, the borough or city council supplier lists, local sponsorship pages. These are the UK equivalents of the "local prominence" signal and they are largely uncontested.
- Accreditation relevant to the trade. Gas Safe, NICEIC, TrustMark, FENSA, CHAS. In trades especially, these function as an eligibility filter before anything else is considered.
City-level nuance worth knowing
Britain's large cities behave differently from each other in ways that affect how you should write about your coverage.
- London is really thirty-odd markets. "London" as a service area is so broad it carries little meaning; borough and neighbourhood naming ("Islington", "Croydon", "Canary Wharf") is what makes an answer specific enough to be given.
- Manchester and Birmingham have strong metropolitan identities and heavily used regional business press, so local editorial corroboration is unusually available and unusually effective.
- Smaller cities and towns are low-competition in AI answers to a degree that surprises people. If you are the only business in your category with a clear entity record and substantive reviews in a town of 80,000, you will frequently be the only name returned.
Write the pages an assistant can quote
A service-area page that says "we provide plumbing services in Leeds and the surrounding area" is unquotable. One that states which postcodes you cover, your call-out hours, whether you charge for a callout, your accreditation number and how quickly you attend an emergency gives a model five separate facts it can use in an answer. Specificity is the whole technique — see our content writing guide for the general form.
A practical order of work
| Step | Action | Why it matters |
|---|---|---|
| 1 | Reconcile NAP everywhere it appears | Removes the contradictions that make a model hedge. |
| 2 | Publish explicit coverage and pricing detail | Gives the model quotable facts instead of adjectives. |
| 3 | Ask customers for descriptive reviews | Supplies independent language about what you actually do. |
| 4 | Claim local and accreditation listings | Adds verifiable corroboration in UK sources. |
| 5 | Re-run your local prompts monthly | The only way to see whether any of it worked. |
Frequently asked questions
Do we need a separate page for every town?
Only where you genuinely operate differently — different coverage, pricing, team or response time. Near-identical town pages read as thin to search engines and give an assistant nothing distinct to say.
Does our Google Business Profile still matter?
Very much. It remains the most authoritative structured record of a local business, and assistants draw on it heavily. It is necessary but no longer sufficient.
How do we track local AI visibility?
Fix a list of the questions a customer in your area would ask, run them monthly against each engine, and record who gets named. Same method as national tracking, scoped to your patch.
Related reading: AI search optimisation in the UK and how to get a UK business cited. For a done-for-you programme, see our UK AI SEO service.