// AI Visibility Audit · for IR35 & contractor accountants
A real contractor-accountant audit:
"[Firm]: specialist accountants for limited-company contractors and IR35"
When a contractor asks AI for an IR35 accountant, does this firm come up, and what would move the needle? Below is the audit we actually ran, anonymised. It's the exact deliverable a founding-cohort firm receives.
// AI-Visibility Foundations Score · real tool output
80/100
🟠 Amber · Partial foundations
80 / 100 on the fixable foundations the AIs read: a genuinely strong base. Two gaps cost the points: no FAQPage markup (−12) and no llms.txt (−8), both cheap, high-leverage fixes. The two included fixes take this firm to about 100 (Green). From a real foundations_check run, 4 Jun 2026.
Your AI-Visibility Foundations Score measures the fixable technical foundations that let AI assistants read and classify your website. It is not a prediction or guarantee that any AI tool will name, rank or recommend you.
// How to read this audit
Two distinct layers, kept honestly separate. The foundations scorecard (Section 4) is real: every mark is genuine output from our automated check against the firm's live homepage on 4 Jun 2026, and the AI-Visibility Foundations Score of 80/100 (Amber) is the tool's real result. The AI-engine visibility matrix (Section 2) is an illustrative layout, not measured data: for this anonymised sample we ran the foundations check but not the engine reads, so that table shows the format a live audit fills from real engine runs. Never read its cells as results. We report the current state of third-party AI tools and best-practice fixes; we don't control those tools and don't guarantee rankings, recommendations, traffic or revenue.
// Before you read on
An honest note
This is a real audit. Yours covers your firm specifically.
A real audit, anonymised. Permission isn't required because nothing identifies the firm.
The audit below was genuinely run against the live homepage of an independent UK IR35 / contractor-specialist accountancy firm. We've anonymised it with permission-not-required, because every identifier (the firm name, its domain, phone, postcode, street and city) has been stripped and replaced with a neutral token ([Firm], [City], and so on). Nothing here can identify the real firm.
We show it so you can see exactly what a Renownly audit looks like before you commit. Your audit covers your firm specifically: your real homepage, your real foundations score, and, in a live paid audit, your real per-engine results in place of the illustrative matrix below.
// Section 1
What we tested
Eight high-intent contractor queries, plus a branded accuracy check.
We test the questions a limited-company contractor would actually ask (IR35, inside vs outside, umbrella vs limited, "near me"), not vanity terms. The queries below are the real set, with the firm's city held as [City] to anonymise it. For the same foundations check run across a contractor cohort, see our 2026 UK Contractor-Accountant AI-Visibility Report.
- Q1"Best contractor and IR35 accounting in [City]"
- Q2"Best accountant for IT contractors in [City]"
- Q3"Best accountant for IR35 contractors"
- Q4"Umbrella vs limited company accountant"
- Q5"Limited company contractor accountant [City]"
- + 3 further unbranded discovery variants ("recommended…", "near me", "top rated…"), and 1 branded accuracy check (the firm's own name)
Methodology note
- Engines (the full audit)
- ChatGPT, Perplexity, Gemini and Claude. A full paid audit reads each one and replaces the illustrative matrix with dated per-engine evidence.
- Runs
- A full audit runs each query five times per engine, reports the consensus and notes the variance. Blank runs/variance cells are the QA tripwire for an under-run audit.
- Foundations
- Section 4 is a real automated check of the firm's live homepage on 4 Jun 2026: public data only, point-in-time.
- Dated & honest
- AI answers are non-deterministic and change without notice; web results are US-weighted, so UK "near me" / AI-Overview conditions are approximated. No rankings, traffic or revenue are guaranteed.
// Section 2
AI visibility results
The visibility matrix.
This is the highest-value page of a live audit: per engine, whether the firm is named, who is named instead, and which sources the engine leaned on. Below it is shown as the layout only, so read the notice first.
// ⚠ Illustrative layout · not measured data
The table below shows the exact format a real audit fills from live engine runs. For this anonymised sample we ran the foundations check but not the engine reads, so the cells below are an illustration of the layout, not measurements of this firm. Do not read any value here as a real result. In a live audit, every cell is replaced with dated evidence from five runs per engine, with the runs and variance sub-values filled. A live audit fills this from real engine runs.
| Query | ChatGPT | Perplexity | Gemini | AI Overviews | Named instead / sources |
|---|---|---|---|---|---|
| Best contractor / IR35 accounting in [City] | Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
[illustrative: national contractor specialists + directories, e.g. ContractorUK, Unbiased, FreeAgent] |
| Best accountant for IT contractors in [City] | Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
[illustrative] |
| Best accountant for IR35 contractors | Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
[illustrative] |
| Umbrella vs limited company accountant | Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
[illustrative] |
| Limited company contractor accountant [City] | Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
Not named[illustrative] |
[illustrative] |
| Branded: "tell me about [Firm]" | Named (accurate)[illustrative] |
Named (accurate)[illustrative] |
Named (accurate)[illustrative] |
n/a (branded)[illustrative] |
n/a (own brand) · [illustrative: own site, Companies House, directory profile] |
Named: firm recommended / described accurately
Not named: a rival or directory appeared instead
[illustrative] = layout placeholder; a live audit fills these from five runs per engine with variance
How to read this once it's real: the illustrative pattern above, strong on the branded row and absent on unbranded discovery rows, is the typical shape for a firm with good brand recall but weak discovery presence. The live run confirms or corrects it. It is a working hypothesis here, not a measured result.
// Section 3
The gap
Likely strong on its own name, weak on contractor discovery.
Grounded in the real foundations findings (Section 4), the likely pattern is a clean split: good branded visibility, weak unbranded discovery. When a contractor already knows the name, an assistant finds [Firm] and describes its IR35 specialism correctly. When they ask an open question like "best accountant for IR35 contractors", the named slots tend to go to aggregators and national contractor specialists. This is a hypothesis, not a measured engine result; a live audit confirms it with five runs per engine.
// where the enquiries go
Directories win the unbranded queries
Unbranded "best… " answers repeatedly route through contractor-accountancy aggregators and comparison sites: ContractorUK's accountant directory, Unbiased, FreeAgent's "find an accountant". A thin presence on the contractor-specific ones is a recurring reason a competent firm stays invisible on discovery queries.
// what winners have
Question-shaped pages for the IR35 buyer
The firms that surface rank with pages titled like the question: "Inside vs outside IR35", "Umbrella vs limited company". [Firm]'s contractor content is more brand-led than query-led, so an assistant has less to match to those exact asks.
Branded recall is the firm's strength
On a branded query an assistant finds [Firm] and describes the contractor / IR35 specialism correctly. That's a genuine asset: the reputation is there, it just isn't reachable from open, unbranded questions yet.
The biggest gaps are cheap to close
The two amber foundations findings, no FAQPage markup and no llms.txt, are exactly the directly-extractable Q&A signal the winners have. Both are low-effort, high-leverage fixes (see Section 5).
// Section 4
Foundations scorecard · real
The signals AIs read, line by line, from a real run.
Auto-filled from a real automated check against [Firm]'s live homepage, 4 Jun 2026. Public data only, point-in-time. ✓ = good · ! = needs attention · ✗ = problem. AI-Visibility Foundations Score: 80/100 (Amber): two gaps cost the points, no FAQPage (−12) and no llms.txt (−8). Lines tagged tool are direct tool output.
- Homepage readability : goodtoolPage returned readable content (HTTP 200); title and meta description present; roughly 8.5k characters of visible text.
- schema.org markup : needs attentiontoolPresent: 2 JSON-LD blocks (AccountingService, WebSite), a solid foundation many firms lack, but no FAQPage. Missing the easiest way to feed AI directly extractable Q&A.
- NAP / location consistency : goodtoolVisible name/address/phone matches the firm's own JSON-LD (phone and postcode agree); the target city is consistent with the address region.
- AI-crawler access (robots.txt / WAF) : goodtoolNo AI crawler is meaningfully blocked; robots.txt carries only wildcard rules (no AI-specific Disallow) and a declared sitemap; no WAF / Cloudflare interstitial served to automated clients.
- llms.txt : absenttoolAbsent (HTTP 404). Emerging and optional, a cheap edge while most rivals also lack it, worth adding now.
- XML sitemap : goodtoolNot at /sitemap.xml, but 1 sitemap is declared in robots.txt, so crawlers can still discover the pages.
- NAP signals on site : goodtooltel: 1 · email: 0 (no mailto: link found) · UK postcode: 1. Contact and about pages reachable (HTTP 200).
AI-Visibility Foundations Score: 80/100 (Amber · Partial foundations). The two items that cost the points, no FAQPage (−12) and no llms.txt (−8), are the headline foundations gaps, and both are cheap, high-leverage fixes (Section 5). The two included fixes take this firm to about 100 (Green). Your AI-Visibility Foundations Score measures the fixable technical foundations that let AI assistants read and classify your website; it is not a prediction or guarantee that any AI tool will name, rank or recommend you.
Manual / off-site checks, confirmed by hand in a live audit
The automated tool can't see these; a live audit checks them by hand. Here they're marked ? confirm because this anonymised sample doesn't expose off-site profiles.
- Google Business Profile: claimed & complete? confirm
- ICAEW / ACCA / AAT "find an accountant": confirm
- Trustpilot / Google reviews: volume & recency
- Unbiased: profile present?
- ContractorUK accountant directory niche · high-intent
- FreeAgent "find an accountant": confirm
- LinkedIn company page: active?
- Companies House: active, correct registered address
The contractor-specific directories (ContractorUK, FreeAgent, Unbiased) are the recurring signal that wins unbranded IR35 queries, which is why a live audit confirms presence on each one by hand.
// Section 5
Your three fixes
Prioritised, firm-specific, tied to the real findings.
Quick win, medium and strategic. Each in plain English with the "why", and each tied directly to a real Section 4 finding, ready to hand to your web person.
// ~1–2 hours · in-house / web person
Fix 1 · Quick win
Publish a /llms.txt and add FAQPage JSON-LD to your key IR35 pages.
Your site already serves clean, readable content with AccountingService schema (good), but the tool found no FAQPage markup and no llms.txt (both real Section 4 findings). Add an FAQ block (marked up as FAQPage) answering the exact questions contractors ask ("Am I inside or outside IR35?", "Umbrella vs limited company: which is better for me?", "How much does a contractor accountant cost?") on the relevant service pages, and ship a one-page /llms.txt listing those key pages. Why: FAQPage markup is one of the most reliable ways to get an assistant to lift a Q&A answer that names your firm; llms.txt is a low-effort edge while most rivals lack it. Both are direct fixes to the two amber rows in your scorecard. Want to do it yourself? Our copy-paste FAQ schema guide and AccountingService schema guide have ready-to-fill blocks.
// ~half a day to a few days · content
Fix 2 · Medium
Build question-shaped pages for the unbranded IR35 buyer.
Your contractor content is brand-led; the firms that surface on "best accountant for IR35 contractors" and "umbrella vs limited company accountant" rank with pages titled like the question. Create two or three pages that answer those literal queries for limited-company contractors (inside vs outside IR35, umbrella vs Ltd, contractor accountancy cost), each with the FAQPage markup from Fix 1. Why: the Section 3 gap is that discovery queries go to aggregators and rivals; question-shaped, well-marked-up pages are how a specialist firm earns its way back into those answers.
// ~2–3 weeks · content + outreach
Fix 3 · Strategic
Claim and complete the contractor-specific directories assistants cite.
Unbranded "best… " answers route through aggregators. Claim and complete strong, review-rich profiles on the contractor-specific ones (ContractorUK accountant directory, FreeAgent "find an accountant", Unbiased), not just the generic ones. Why: assistants lean on these aggregators for unbranded discovery; being inside the contractor directories (where your IR35 specialism is the differentiator) is how you start appearing on the exact queries your ideal client asks. Confirm current presence via the off-site checklist in Section 4 first.
What's your Foundations Score?
This firm scored 80/100. Check your own AI-Visibility Foundations Score free in under a minute: the same real number, scored on your live homepage. Then claim a place in the founding cohort for the full audit, with two quick-win fixes done for you.
Based on a real Renownly audit, IR35 firm anonymised. The findings above are genuine, from an audit we actually ran on an independent UK IR35 / contractor-specialist accountancy firm. We've stripped every identifier (name, domain, phone, postcode, street, city) and replaced each with a neutral token, so the firm can't be identified; anonymisation is permission-not-required because nothing here points to the subject. The Section 4 foundations scorecard is real tool output (AI-Visibility Foundations Score 80/100, Amber, 4 Jun 2026); the Section 2 AI-visibility matrix is an illustrative layout, not measured data, and a live audit fills it from real per-engine runs (≥3 per engine, with variance). The Section 3 gap narrative is a working hypothesis grounded in the real foundations findings, not a measured engine result. Renownly reports the current state of third-party AI tools and gives best-practice recommendations. We do not control those tools and do not guarantee rankings, recommendations, traffic or revenue. AI answers are non-deterministic and change over time; all findings are dated at the time of testing. Foundations findings are drawn from public data only. AI visibility is not the same as SEO ranking.