W-03 · North London repair shop case study
Customers now ask ChatGPT and Google's AI which repair shop to trust. We made sure the answer was our client.
The short answer
AiS LABZ audited the shop the way customers now search: by asking AI assistants. We then rebuilt its Google Business Profile field by field and shipped schema and llms.txt. On 20 September 2026 the shop was named in 4 of 4 Google AI Mode answers and named first by ChatGPT and Gemini.
What was in the way?
- One Business Profile category and 26 services with no descriptions.
- No profile posts for six months, 12 unanswered reviews, an empty Q&A.
- Booking link pointed at the homepage, not the booking page.
- 32 blog posts with no meta descriptions and none mentioning its neighbourhood.
- A directory listed the wrong Sunday hours, and an AI assistant repeated it.
What changed?
- Named in 4 of 4 Google AI Mode answers, and cited in 3.
- Named first by ChatGPT and by Gemini for local repair prompts.
- 1,066 customer interactions on the Business Profile.
- 24 of 24 services now carry full descriptions.
- 0 of 45 posts with broken embedded images after the blog repair.
How did we do it?
The same four-phase method on every engagement: diagnose, model, build, compound.
Diagnose
17 real customer prompts across Perplexity, Google AI Mode, ChatGPT and Gemini, plus a crawl of 13 pages and 32 posts.
Model
Rank every fix by how directly it changes what an AI assistant says. Open standards only: schema.org, robots.txt, llms.txt, accurate listings.
Build
Five new profile categories, a 739-character description, 24 services described, the booking link repointed, the primary number unified, eight posts scheduled and review replies written. Schema and llms.txt shipped on the site.
Compound
132 link placements logged since March 2026 and a monthly AI-visibility recheck.
Who did we work with?
What happens next?
Q&A seeding through Maps, directory corrections, and the 20 October AI-visibility recheck.
Common questions
Not by trickery. AI assistants repeat what trustworthy sources say. We make those sources accurate: your site's schema, your Business Profile, and the directories they read.
Because that is how customers now shortlist a local service. If the assistant gets your hours wrong, you lose the visit before search even starts.
The same fixed prompt set, run monthly, recording whether the shop is named, where, and whether its own site is cited.
Your number next