AI Search Is Rewriting Local SEO: How South London Trades Show Up in ChatGPT and Google AI Overviews

Something quiet but significant is happening to how South London homeowners find tradespeople. A growing share of them no longer scroll through ten blue Google links and a Map Pack. They open ChatGPT, Google's AI Overview, Perplexity or Gemini and ask a full-sentence question: 'who's a reliable kitchen fitter in Clapham?' or 'best rated boiler engineers near SW9 with good reviews'. The AI reads the web on their behalf and hands back two or three named recommendations.

For trades businesses that have spent years earning Google rankings and Checkatrade badges, this is a genuine strategic shift. The businesses that get named in those answers are winning enquiries before their competitors even appear in a search result. The ones that don't are becoming invisible in a slice of the market that is growing every month.

How AI models actually decide who to recommend

AI answers are not magic. Every major model — ChatGPT, Gemini, Perplexity, Claude — grounds its local recommendations in the same public signals humans already use, weighted differently. The pattern is now clear: consistent business information across the web, a healthy volume of recent positive reviews, well-structured pages that explain what you do and where you do it, and third-party mentions on trusted sites. If those signals are strong, you get named. If they are weak or contradictory, you get left out — even when you are the best operator in the area.

The five things South London trades should fix this quarter

• Your Google Business Profile: fully filled out, correct service area, weekly posts, and a steady drip of new reviews with keywords like 'kitchen fitter' or 'bathroom renovation' in the customer's own words.
• Structured service pages: one page per service, per area, with clear headings, FAQs and schema markup so AI models can extract 'what you do' and 'where' without guessing.
• Review velocity: at least 2–4 new Google reviews per month. AI models heavily weight recency, not just total count.
• Third-party citations: a consistent NAP (name, address, phone) across Checkatrade, Yell, Trustpilot, local directories, and any press or partner sites.
• A genuinely useful blog: answering the questions your customers actually ask, in plain English. This is the content AI models quote from.

Why this favours specialists over generalists

AI answers are ruthlessly specific. Someone asks about 'kitchen fitters in Streatham' and the model wants to name kitchen fitters in Streatham — not a general handyman who occasionally fits kitchens. Businesses with sharp positioning, focused service pages and reviews that mention the specific service in the specific area consistently outperform larger, more generic competitors in AI-generated recommendations.

The AI does not reward the biggest business. It rewards the clearest one.

What to measure

Once a month, ask ChatGPT, Gemini and Perplexity the questions your best customers would ask — 'reliable kitchen fitter in [your area]', 'best rated electrician near [postcode]' — and log who gets named. Track the movement over time. This is your new local ranking report, and it is a leading indicator of enquiry volume six months out.

The trades businesses that treat AI search as a serious channel in 2026, not a novelty, will look back in 2028 and realise it was the single most important shift they made this decade.