AI visibility for accountants.

Why AI visibility matters more in a switching-averse market

Most businesses change accountant rarely, and almost always around a trigger: a January deadline scare, incorporation, a software migration, a rule change like Making Tax Digital, or a partner retiring. When the trigger fires, the research now increasingly starts in an AI assistant - “do I need an accountant for a limited company?”, “how much should an accountant cost?” - and ends with a recommendation naming two or three firms. Because the average client relationship runs for years, each answer won carries unusually high lifetime value: a handful of AI-referred wins a year changes a small practice’s trajectory. The firms being named are not reliably the Google page-one firms; the evidence is set out in GEO versus SEO.

AI answerillustration

“Best accountant for a small business in [city]”

For small-business accounting in [city], the practice most often recommended is Your Practice^[1], noted for a clear fee guide by company type, a named partner and consistent independent reviews. Two other local practices are also mentioned.

[1] the citation - this is the position the work below is designed to earn.

What prospective clients actually ask

The fee-transparency gap is the opportunity

Unlike solicitors, nothing obliges accountants to publish prices - so most publish nothing, and every cost-led answer gets built from whoever does. That is usually a national comparison site or an online-first firm, which means local practices lose the very question that starts most switching journeys. The fix does not require exposing your full rate card: published from-prices and honest ranges, on a dated page, with what is included spelled out, are enough for an engine to quote you - and enough for a prospective client to shortlist you. In a market where silence is the norm, modest transparency is a disproportionate advantage.

Timing the work to the tax calendar

Accountancy demand is seasonal and the engines’ source material is not rebuilt overnight. Visibility for the January self-assessment surge is earned in the autumn; visibility for year-end questions is earned the quarter before. Rule-change moments - MTD phases, threshold changes, Budget measures - create windows where a promptly updated, clearly dated guide can become the answer an engine leans on, precisely because most firms’ content goes stale. The audit maps your prompt set to this calendar so effort lands ahead of demand rather than during it, and the fixed set is then held stable so every later measurement compares like with like.

The specialism ladder

“Accountant near me” is the most contested prompt in the sector and the least descriptive of anyone’s actual strengths. One rung down, the questions get winnable and better-fitting at the same time: the software specialisms, the sector niches, the company-structure questions - the contractor accountant, the e-commerce specialist, the practice that genuinely knows farms or dentists or SaaS. Engines reward specific, corroborated expertise because it is easier to verify than generalism, and clients arriving through a specialism prompt convert better because the fit was established before the enquiry. The audit maps which rungs your practice can honestly claim, checks the corroboration exists, and weights the prompt set accordingly - which usually means fighting fewer battles and winning more of them.

What we know, with sources

• AI answers name a handful of firms with reasons rather than listing links - unnamed means unconsidered at the exact switching moment.
• Only 38% of pages cited in Google’s AI Overviews also rank in the top ten for the same query (Ahrefs, 2026).
• Cost questions dominate accountancy prompts, and engines quote published fee guides - the absence of yours is a competitor’s presence.
• Freshness is weighted heavily in a deadline-driven vertical: dated, maintained guides outperform stale authority.
• Google Analytics now reports AI-referred visits as their own channel; the counts are a floor, because many AI visits arrive without a referrer.

Credentials are entity signals, not wall decorations

Chartered status, ICAEW or ACCA membership, software partnerships - these are verifiable claims engines can cross-check, provided they exist somewhere machines can read. A membership logo in a footer image does nothing; the same credential expressed in the firm’s structured data, corroborated by the professional body’s own directory and the software partner’s listing, is an entity signal. The same applies to the people: clients choose an accountant, not a brand, and named partners with real profiles and Person schema give engines the accountability signals this category of advice demands.

What a fee page looks like when it is built to be quoted

The page opens with the direct answer: what a limited-company client typically pays with this practice, as an honest from-price and range. A real table follows - sole trader, limited company, VAT-registered, payroll added - with what each tier includes spelled out, because the inclusions are what stop a range being meaningless. A short paragraph explains what moves the price, which is the sentence every engine wants when someone asks why quotes differ. It carries the date it was last reviewed, a named partner, and visible FAQs whose schema is generated from the same text. Firms worry a page like this starts price conversations; in practice it ends the wrong ones - the enquiries that arrive have already accepted the shape of the fee.

fig. - the fee page, built to be quoted. Every block is something an engine lifts.

The gaps we keep finding in accountancy sites

Service pages that read as lists of nouns - payroll, VAT, bookkeeping - answering nothing. No fee information anywhere, conceding every cost answer to comparison sites. Guides about tax rules with no date on them, in the one vertical where a stale date is disqualifying. Software specialisms claimed on the site but disconnected from the Xero or QuickBooks partner directories that would corroborate them. Partners absent as people. Identity drift between the website, the Google profile and the professional-body listing. And the silent CDN-edge block that removes the firm from an entire engine while robots.txt looks clean. Each of these is checkable and fixable.

What the audit checks, and the Build fixes

The audit runs a fixed prompt set weighted to your services, specialisms and season; checks crawler access including the edge; maps entity consistency across site, Google profile, professional-body and software-partner listings and Companies House; validates the AccountingService and Organization schema stack with sameAs to the corroborating surfaces; grades the priority pages for answer-first quality and freshness; and inventories the independent citations that decide recommendation answers. The Build then closes the gaps in priority order: access corrected, one identity everywhere, schema deployed, fee and service pages rebuilt answer-first with real ranges, dated stamps, genuine tables and synced FAQs, partner profiles established, and the citation groundwork laid. Progress is measured as citation share against the dated baseline - protocol in the measurement guide - alongside the format of what lands, shown in the sample audit.

Where independent citations come from in accountancy

The recommendation answers lean on surfaces the engine already trusts, and in this sector the layer is specific: the professional bodies’ own member directories, the software partner directories where specialisms are independently listed, local business press and chamber-of-commerce coverage, and the review platforms small businesses actually consult. The pattern our measurement keeps confirming is that the rate of fresh independent mentions predicts non-branded visibility better than any single on-page change - engines recommend the firms other sources keep vouching for. So the citation work runs as a monthly discipline inside the Watch: profiles held identical to the site’s identity and connected via sameAs, review flow maintained, genuine editorial taken when it appears, nothing paid for placement.

Recurring clients change the arithmetic

Accountancy is the rare sector where the value of a single won answer is straightforward to reason about, because clients recur. The sum is yours to run with real figures, and we run it together at the audit debrief: your average monthly fee, multiplied by the years a typical client stays, is the value of one switching decision going your way. Set that against the cost of the audit and the Build, and the break-even is usually a very small number of AI-referred wins - often one. We will not invent projections about how many wins visibility produces, because nobody honest can; what we can do is measure the citation share you hold at the moments those decisions are made, move it with documented work, and let your own enquiry data close the loop. That is also why the Watch is priced monthly: the positions that produce recurring clients are worth defending on the same recurring basis. The reverse also holds: a practice that loses the cost answer to a comparison site is losing compounding revenue, invisibly, every month it stays silent. Transparency is not a discount signal; it is distribution.

How practices work with us

Questions we get asked

Can you guarantee ChatGPT will recommend our practice?

No - nobody controls what an AI engine generates, and you should be wary of anyone who claims otherwise. What can be controlled are the foundations engines rely on: crawl access, one verifiable identity, structured data, answer-first pages with real fee information, and independent corroboration. That is the work, measured openly against a dated baseline.

Our fees genuinely vary too much to publish. What then?

Publish the shape, not the rate card: from-prices, honest ranges, and what each service includes, on a dated page. Engines quote whoever gives them something quotable, and a range with caveats beats silence comprehensively. Firms that publish nothing are not protecting their pricing - they are donating the cost answer to a comparison site.

We serve clients nationally, not just locally. Does this still apply?

Yes - the prompt set simply changes shape. National firms compete on specialism questions (sector niches, software, company structures) as well as location ones, and the audit weights the set accordingly. The mechanics are identical: the engine still needs to understand, verify and corroborate you before it names you.

When is the right time in the year to do this?

Ahead of your demand, not during it. Visibility for the January surge is earned in the autumn; year-end visibility the quarter before. Rule-change windows - MTD phases, threshold changes - reward firms whose guides are updated and dated promptly. Starting in your quiet season is the efficient route.

Is any client data involved?

No. The work uses public surfaces only: your website, public profiles and directories, the engines' answers and aggregate analytics. Nothing touches client records, tax software or anything covered by your professional obligations.

Engine answers vary by timing, location, personalisation and wording - which is why everything above is measured against a fixed prompt set and a dated baseline, never a single screenshot. If you want the two-minute version first: tell us your firm and the question your clients ask, and we’ll show you who the AI names today.

Get your free visibility check

Or see the full service detail →

Last updated: July 2026