Across high-intent UK business-insurance queries run across Claude, GPT, Gemini and Perplexity, Admiral appears in 0.58% of competitor rankings. Simply Business appears in 4.85%, AXA in 2.88%. Admiral sits 37th in the competitor table, behind regional brokers we would expect Admiral to outrank. When Admiral does surface, sentiment is 14 positive against 1 negative. The data describes a presence problem rather than a perception problem, and presence problems are addressable with content and structure.
Top 10 by share of LLM rankings, plus Admiral's actual position. Sentiment is positive / neutral / negative across the sample. Provider breadth = how many of the 4 LLMs surface the brand.
| # | Brand | Mentions | Share | Avg rank | Avg vis. | Breadth | + / · / − |
|---|---|---|---|---|---|---|---|
| 1 | Simply Business | 143 | 4.85% | 3.5 | 63 | 4/4 | 112 / 12 / 0 |
| 2 | Hiscox | 115 | 3.90% | 4.2 | 54 | 4/4 | 89 / 9 / 0 |
| 3 | Direct Line for Business | 92 | 3.12% | 6.0 | 36 | 4/4 | 63 / 15 / 1 |
| 4 | AXA | 85 | 2.88% | 5.2 | 41 | 4/4 | 65 / 6 / 1 |
| 5 | Markel Direct | 80 | 2.71% | 5.6 | 38 | 4/4 | 59 / 11 / 1 |
| 6 | AXA UK | 66 | 2.24% | 4.4 | 51 | 4/4 | 54 / 5 / 1 |
| 7 | Rhino Trade Insurance | 57 | 1.93% | 3.7 | 61 | 4/4 | 36 / 8 / 0 |
| 8 | Tradesman Saver | 54 | 1.83% | 5.9 | 34 | 4/4 | 37 / 4 / 0 |
| 9 | Aviva | 51 | 1.73% | 6.3 | 32 | 4/4 | 44 / 1 / 0 |
| 10 | PolicyBee | 45 | 1.53% | 4.0 | 52 | 4/4 | 40 / 1 / 1 |
| ⋯ rows 11 to 36 omitted ⋯ | |||||||
| 37 | Admiral Business | 17 | 0.58% | 4.8 | 53 | 4/4 | 14 / 2 / 1 |
6 appearances on a single query, averaging rank 1.67. Wins rank #1 on Perplexity, Anthropic and Gemini. The reasons LLMs give: visible pricing (£7.08/month), Cardiff/South Wales focus, designer specialism.
Cardiff tradespeople (rank 5 avg), Cardiff builders (rank 3), Birmingham designers (rank 3), London designers (rank 6.7), Leeds designers (rank 6).
Leeds tradespeople (447 rankings), London contractors (310), Leeds builders (221), London builders (184), Manchester contractors (174), Manchester builders (127), Manchester designers (94), and 6 more.
Anthropic is the surface most likely to mention Admiral today (9 of 17). Across the others, Admiral lands two or three times each. The pattern suggests Admiral's content is being read by some retrieval pipelines and missed by others, rather than being absent everywhere.
2,101 free-text reasons extracted and tagged from LLM responses. The pattern is consistent across all four providers and is the most actionable part of this audit.
"Specialist insurer for…", "30k+ customers", "decades of experience". Wins go to sites with deep per-niche landing pages, named expert authors, and topic-cluster depth.
"4.5/5 Trustpilot", "9/10 Feefo", awards. Structured-data review badges, embedded Trustpilot widgets, and named case studies are what get cited back.
"From £7.32/month", "public liability from £85/year". Concrete £ numbers in body copy are quoted back verbatim. The £7.08 number wins Admiral Cardiff.
LLMs decide what to cite via two distinct loops on very different timescales. Strategy needs to address both, but the fast loop is where Admiral can move in the next four months.
LLMs are trained on a snapshot up to a fixed cutoff (currently 6–18 months behind today). Content published this week cannot influence what a base model "knows" until the next training run.
For Admiral: content shipped in 2026 populates the 2027 training corpus. Pays back over the long tail.
Increasingly dominant. ChatGPT Search, Perplexity, Claude with web access, Gemini and Google AI Overviews pull live web results at inference time. Behaves like classic SEO: get indexed, signals accumulate, retrieval surfaces your URLs.
For Admiral: content live in May 2026 can be cited in answers in June 2026. Cardiff is the proof point already in-sample.
First Googlebot indexation. First AI-bot crawls (visible in Bot Insights). First long-tail rankings. Early GSC impressions.
First repeatable AI-citation hits. Daily variance still ±30%. Sample size matters.
Citation share stabilises per provider. Cross-provider breadth (3+ LLMs on same query) is the right success metric.
Each path is a defensible answer to the data above. The trade-offs:
Start with the Path B 4-month pilot. The Cardiff signal is a working proof that Admiral can rank when depth, geography and niche align, and the main site does not have the URL surface area to test that hypothesis in three different shapes simultaneously. If Phase 1 lands at the month-4 gate, Path C is the natural escalation: an embedded venture studio inside Admiral Pioneer scaling the motion across new categories at platform speed. Path C can also be committed to up front, only sensible if Pioneer's strategic prize is the operating model itself rather than any one microsite win.
Phase 3 integration patterns (copy, hub-and-spoke, 301-redirect) mean Admiral keeps the playbook and content regardless. The Confused.com precedent is the strongest counter to the concern.
Freelance procurement is days, external vendor is weeks. If procurement is impossible in any reasonable window, Path A is the right answer by default. Shipped beats stalled.
A 2025 footer, missing schema, no llms.txt, and a 700-word body on key category pages. None of this fixes itself. Should be scoped regardless of path.
We use cookies to understand how visitors use our site and improve your experience. Learn more.