Where Waggel stands in AI answers, and the plan to change it

01 · Where Waggel stands

Already trusted as a source. Not yet in the consideration set

In the UK AI Overviews corpus, waggel.co.uk is cited in 5,903 answers, genuinely competitive with ManyPets (6,713) and within reach of Petplan (12,476). The composition is the issue, not the volume: those citations come overwhelmingly from Waggel's breed guides (the questions citing waggel.co.uk most are "cavapoo", "labrador" and "mini dachshund", each 50k+ monthly AI queries), not from insurance-intent questions.

Ranking the corpus's UK pet insurance questions by how often people actually ask them, the highest-volume questions are generic head terms and comparison intents. Those answers are dominated by comparison sites and the incumbent insurers, which is exactly why Petplan (founded 1976) sits at 60/100 while Waggel, whose authority today lives in breed content, appears in 6:

The AI engines already treat waggel.co.uk as a trustworthy source; they just don't yet associate the brand with the buying-intent questions. When Waggel does appear, the framing is already right:

"Best for Young Pets: Waggel, known for low, customizable, and modern pet insurance policies."

What UK pet owners actually ask AI tools

The same top-100 questions, grouped by what they are about. "Monthly AI questions" is how many times per month UK users put questions from that group to AI tools; "include Waggel" is how many of those questions Waggel currently appears in:

02 · The mechanism

AI answers are built from fan-out queries

When an AI engine answers a question, it doesn't search that question once. It issues a fan of specific sub-searches, Google calls this "query fan-out", and synthesises the answer from what those return. Our tracking engine records every sub-search the models issue on every scan. This is one real example from last week's production scans:

The article targets, cleaned and grouped

We extracted 968 unique fan-out queries from three evidence layers: last week's tracking scans, early-July scans, and this morning's direct model probes. Below is the cleaned, deduplicated target list: UK-relevant queries the models genuinely issue, each one usable as an article title or H1. The full 968-row sheet (with theme, frequency and which engines issued each query) is delivered alongside this page:

The engines also hunt authority reviews by name (Which? Best Buy, MoneySavingExpert and Martin Lewis, Forbes Advisor appear across dozens of sub-searches), which is why review-format and comparison-format articles are part of the mix.

03 · Evidence

Sites with near-zero Domain Rating get cited next to Which? and Forbes

These are the sources AI platforms actually cited when answering Waggel's tracked questions over the last 30 days, with each site's Ahrefs Domain Rating alongside. If high DR were required to be cited, the bottom half of this table could not exist: vetcost.co.uk (DR 0.4) and petcoverpro.co.uk (DR 0.0) are cited by 7 or 8 of the 8 platforms we track, at rates comparable to which.co.uk (DR 89) and forbes.com (DR 94). What these low-DR sites share is not authority metrics but topical relevance: they are entirely about pet insurance. This is the empirical basis for the placement criteria in the plan:

We also asked the models directly this morning

As a second, independent check we put 22 UK questions (Waggel's 8 tracked plus 14 intent questions) to ChatGPT, Claude and Gemini with web search on, UK locale, twice each: 132 answers. Where the question already involves Waggel, the models mention it in 60% of answers and cite waggel.co.uk in 48%. On non-branded intent questions: 25% mention, 17% cited. The split is stark:

Once Waggel is in the consideration set, the models say the right things about it, and cite waggel.co.uk. The entire problem, and the entire opportunity, is the non-branded intent queries on the right.

04 · Plan of action

Blanket the fan-out queries Waggel is absent from

Each fan-out query is a concrete article target: a piece whose title and structure match the query the engines actually issue, published on a topically-relevant site, with Waggel present as an unlinked brand mention among the honest options. This is a volume play, not a metrics play, mass coverage across many relevant publications seeds the Common Crawl dataset, which AI engines rely on twice: for foundational training (the probability of "Waggel" appearing next to "pet insurance") and for retrieval-augmented generation (live web search). The 968-query master list is the targeting source; the top-100 question gaps set the priority order.

Priority order

Sequenced by monthly AI search volume × Waggel absence, the order below is the build order:

1. Pre-existing & chronic conditions, the single biggest clean gap16 top-100 questions · 62,900 vol/mo · Waggel in 0 · 138 fan-out queries mapped
2. Dog insurance cluster (incl. lifetime cover)36 top-100 questions · 187,500 vol/mo · Waggel in 1 · 97 lifetime fan-out queries
3. Senior / older pets, heavy branded fan-out to competitors today102 fan-out queries · engines verify age limits brand-by-brand · Waggel absent
4. Dental & specific coverage questions28 fan-out queries · "does pet insurance cover dental UK" issued by every engine
5. Breed × insurance crossovers, leverage authority Waggel already hasengines already cite Waggel's breed guides; extend that trust to breed-insurance intent
6. Vet access / perks & claims speed, Waggel's proven differentiatorsWaggel already wins these when present (100% mention in probes), widen the surface

How we'll know it's working

The measurement is already running. Weekly scans across all 8 AI surfaces (the dashboard Ryan has access to) track presence, share of voice and citations per question, the same instrumentation that produced every number on this page. Independently, the DataForSEO UK corpus provides a monthly external benchmark: mentions, new/lost citations and question-level presence, measured from real AI Overview responses we don't control.

Honest expectations on timing: retrieval-driven surfaces (Google AI Overviews / AI Mode, Perplexity, search-backed ChatGPT answers) can reflect new content within weeks of crawl + indexing. Brand-entity effects compound over months as the seeded pages enter Common Crawl and flow into the next model training cycles. The weekly scans make progress, or the lack of it, visible either way, and the target list adapts as citation data comes in.