A UK sportsbook — from invisible to cited in ChatGPT, in four months.

The challenge

The client was a challenger UK sportsbook — well-established in the market, with strong organic search visibility on traditional Google results, but completely absent from AI-generated responses. When a potential customer asked ChatGPT or Perplexity "what are the best UK betting sites," the client's brand didn't appear. Competitors with weaker traditional SEO profiles were being cited consistently.

The brief was clear: build AI search visibility on the highest-intent commercial queries in one of the UK's most competitive verticals.

Our approach

We started with a citation audit across ChatGPT, Perplexity, Gemini, and Google AI Overviews for 40 target queries — "best UK betting sites," "best sportsbook for football," and similar high-commercial-intent terms. The audit established a baseline: zero citations across all four platforms on all target queries.

The diagnostic pointed to three root causes:

Entity footprint weakness. The brand's entity data was incomplete and inconsistent across the sources AI systems use to understand it — Wikipedia, Wikidata, industry directories, press coverage, structured data. LLMs had a weak model of what the brand was and whether it was trustworthy enough to cite.

Content architecture not built for direct answers. The site's content was written for traditional SEO — good keyword coverage, reasonable authority signals — but not structured to answer the specific questions AI systems were being asked. No clear, quotable answers to "why should I use this sportsbook?"

Co-citation signals absent. The brands appearing in AI answers had been consistently mentioned alongside trusted third-party sources — review sites, comparison pages, authoritative press. The client's brand had thin co-citation coverage outside their own properties.

The work

We ran a four-phase programme over six months.

Phase 1 — Entity repair. We audited every major source that feeds LLM training data and identified gaps and inconsistencies. We submitted corrections and additions to Wikidata and structured data sources, worked with the client's PR team to seed consistent brand descriptions across authoritative press, and implemented comprehensive Organisation and WebSite schema across the site.

Phase 2 — Content restructuring. We identified the 12 pages most likely to be cited for target queries and restructured them around direct-answer formats. Each page now opens with a clear, quotable answer to the primary query it targets, followed by substantiated supporting content. We also created a set of net-new FAQ-style pages targeting the exact question forms AI systems were being asked.

Phase 3 — Co-citation seeding. Working with the client's existing PR and affiliate relationships, we placed consistent brand mentions alongside already-cited competitors on high-authority third-party sources — comparison sites, betting guides, financial press. This wasn't link building in the traditional sense; the goal was co-occurrence signals, not PageRank.

Phase 4 — Monitoring and iteration. From month one, we ran daily citation monitoring across all four AI platforms using our own platform. This let us see which content changes correlated with citation appearances and refine the programme accordingly.

---

The outcome

By month four, the client was appearing in 40% of monitored queries across ChatGPT and Perplexity. By month six, that had reached 52% on Perplexity and 38% on ChatGPT.

The most significant move came after Phase 2 — the content restructuring correlated with the first consistent citations appearing, which confirmed our hypothesis that the entity work was a necessary precondition, but direct-answer content was the trigger.

"We'd been told AI search visibility was something you couldn't really control. This programme proved otherwise. The monitoring showed us exactly what was working and when."

The client has since brought the citation monitoring into their standard reporting suite, and the content brief framework we established is now part of their ongoing content production process.

---

Related disciplines

This engagement drew on AI Search, content strategy, structured data implementation, and PR coordination. It is a good example of the kind of work that sits between traditional SEO and AI Search — the entity and co-citation layer that connects both disciplines.