AI Visibility · 24 June 2026

By Eddie Eastwood, Founder

AEO for Accountants & Bookkeepers: Winning AI Referrals

A sole trader going limited, a landlord wrestling with Making Tax Digital, a growing business that has finally outgrown its spreadsheet: each one needs an accountant, and not one of them is shopping for the same accountant. The sole trader wants somebody fluent in owner-managed companies. The landlord wants property tax. The growing business wants whoever already lives inside their bookkeeping software. Three buyers, three different specialists, one job title.

What every one of them is doing, whether they would describe it this way or not, is narrowing. They begin with the vague category "an accountant" and they filter, quickly, down to a particular fit: my sector, my service, my software. And the place that filtering now happens has changed. Rather than open Google and skim a list of firms, a growing share of these buyers put the question to an AI, "what kind of accountant do I need once I incorporate?", or read the AI summary that now sits above the results before they look at a single firm.

Whichever practice the AI lands on once that narrowing settles is the one that gets the call. Answer Engine Optimisation (AEO) is the work of being that practice: the accountant or bookkeeper an AI tool cites and recommends when a business owner asks it for the right fit. This guide is built around the way an accounting client actually decides, because choosing who handles your money is a researched, qualification-led choice, not an impulse, and AEO for a practice has to be shaped to that. No jargon, no enterprise tooling.

The general model behind this lives in our complete guide to AEO for local businesses; this page is the version for a practice whose clients choose on specialism.

The buyer is not searching for "an accountant", they are searching for a fit

The single most useful thing to understand about an accounting prospect is that "find me an accountant" is never the real query. It is the opening of a three-part qualification, and the AI answer is increasingly where that qualification gets resolved.

The buyer narrows along three axes, usually at once:

• Sector. Contractor, landlord, e-commerce seller, restaurant, freelancer, startup. "An accountant who gets my line of work" is the first filter most owners apply, because they want someone who already knows their reliefs, their quirks and their deadlines.
• Service. Limited-company accounts, self-assessment, VAT, payroll, Making Tax Digital, management accounts, tax planning. A landlord and a payroll-heavy employer want very different things from the same word "accountant".
• Software. Xero, QuickBooks, Sage, FreeAgent. A business already running its books on one platform overwhelmingly wants an adviser who works in the same one, and treats that as a near-deciding factor.

A practice that reads as a clear answer on all three axes is the practice that survives the filtering. A practice that presents itself as a generalist "for all your accounting needs" reads as an answer to none of them.

This decision has moved into AI faster than most firms realise, and three measured shifts explain why it matters to act now.

People meet an AI answer in the course of an ordinary week of searching, and AI-generated summaries now sit at the top of a large share of everyday Google results. For a query like "accountant for a limited company near me", that summary now sits where the list of firms used to be.

When that summary shows up, the onward click mostly does not. An Amsive analysis of 700,000 keywords across finance and four other sectors found that pages losing ground to an AI Overview shed 15.49% of their click-through rate on average, with non-branded queries down 19.98% (Amsive, 2025)^[3]. A firm that is not inside the answer rarely gets a second look further down the page.

And buyers are now willing to act on what the answer recommends. BrightLocal's Local Consumer Review Survey 2026 put the share of consumers using ChatGPT and similar AI tools for local business recommendations at 45%, up from 6% a year earlier, which is now the third most popular way to find a local business; 40% said they trust AI tools to recommend businesses, and 42% trust them as much as traditional reviews (BrightLocal, Local Consumer Review Survey 2026)^[1].

Stack those up against a researched, trust-led purchase. The owner who is carefully choosing who to hand their accounts to is now running the early stages of that research through an AI, taking fewer of the onward clicks that once led to your website, and increasingly treating the AI's shortlist as a credible one. If a prospect never sees your firm at the qualification stage, you simply do not enter their consideration, and you have no way of knowing it happened.

What survives when AI rewrites the question

There is a worry worth addressing head-on, because it is the reason many independent firms assume AEO is a game only the big online platforms can win. When the engine takes a question, it does not search for the words the buyer typed. It quietly rewrites and expands them first. The fear is that this rewrite throws away the very things that make you the right answer, the "in Birmingham", the "for contractors", the specialism that is your whole advantage.

Two findings show that fear is largely misplaced, and together they explain why a focused practice is well placed rather than doomed.

The first is how the rewrite works. Google describes its own approach as "query fan-out": it breaks a question into subtopics and issues "a multitude of queries simultaneously on your behalf" (Google, AI in Search, May 2025)^[2]. So "who's a good accountant for a small limited company in Bristol?" does not become one search. It becomes a cluster of hidden ones, running in parallel: local accountants, limited-company specialists, typical fees, qualifications and accreditations, reviews, and the software the firm supports. The engine assembles the reply from whatever it can find against each of those threads. That is good news for a specialist, because it means your fit is being assessed on several specific fronts, not on one generic keyword. The more of those threads your online presence answers plainly, across sector, service, software, fees, qualifications and reviews, the more of them you turn up in, and the more often your name reaches the final answer.

The second finding is what survives the rewrite. When an AI engine reshapes a local query it tends to drop loose adjectives and thin out soft filters, but the location itself almost always survives the rewrite. That fits the independent BrightLocal evidence that buyers settle "near me" and "right for me" questions on location and reviews.

Geography, then, is the filter the engine holds onto when it discards the rest, and a clearly stated specialism is what turns "a local accountant" into "the obvious answer for this client". A national platform has to be the right answer for every sector and every town at once, which is precisely what it cannot be with any precision. You only need to be the clearest, best-documented, best-reviewed answer to "accountant for contractors in [town]" or "bookkeeper for landlords in [area]". That is a far narrower contest, and it rewards the focus an independent firm already has.

The three things that decide a research-led choice

Because the buyer is qualifying you rather than just looking you up, the work that wins an accounting referral is not a generic checklist of marketing tasks. Three things do most of the deciding, and two of them are territory an independent practice can own outright. Everything else in this guide exists to support these three.

Provable credentials, stated identically everywhere

For a profession built on trust, qualifications are not a footnote; they are a primary signal both to a cautious buyer and to an engine deciding whether to put your name in writing. State the professional body membership you genuinely hold, accurately and without inflating it: ICAEW, ACCA, AAT or CIMA for accountants, ICB or IAB for bookkeepers. Then make that claim verifiable. Each of those bodies runs an official member directory, and a listing on one is a high-trust source that an engine can cross-check against your own site. A qualification you assert on your homepage and confirm on the ICAEW or ACCA register reads as a fact; the same qualification asserted only on your own site reads as a claim. Keep the exact wording consistent across your website, your Google profile and every directory, so the engine sees one qualification described one way rather than three near-matches it has to reconcile.

Presence in the software-partner directories

This is the lever most accounting and bookkeeping firms overlook entirely, and it is close to unique to the profession. The accounting software platforms, Xero, QuickBooks, Sage and FreeAgent, each run advisor or partner directories where businesses already on that software go to find a local expert in it. Earning a listing, and partner or certified status where you qualify, does two jobs at once: it is a well-worn route real clients use to find an adviser, and it is a strong, specific signal to an engine answering "accountant who uses Xero in [town]". Because software fit is one of the buyer's three filters, a verified presence in the directory for the platform you specialise in can be more decisive than anything on your own website. Few firms compete here, which is exactly why it is worth the effort.

Sector, service and fee pages that answer the qualifying questions

Your website's job is to settle the questions the buyer is filtering on, in their own words. That means a clear page for each service named the way clients say it ("limited company accounts", "self-assessment tax return", "VAT returns", "payroll", "bookkeeping for small businesses"), describing what is included; a page for each sector you genuinely specialise in, because "accountant for contractors" or "bookkeeper for landlords" is the focused question you can win; and clear fee information, whether a fixed monthly package, a typical range, or how your pricing is structured. A page that answers "how much does an accountant cost for a limited company in [town]?" outright is a page an engine can quote. One that hides the answer behind "trusted advisers to ambitious businesses" is one it cannot use, and a phrase like "bespoke accounting solutions tailored to you" gives the engine nothing concrete to repeat. Write to the question, not to the brochure.

The groundwork those three factors rest on

The deciding factors above only work if the basics underneath them are in place, so treat the following as the foundation rather than the strategy.

A complete Google Business Profile. This is the floor, and a startling number of professional firms never lay it, assuming word of mouth is enough: BrightLocal reports that only about 35% of SMBs have a Google Business Profile at all (BrightLocal, Local Consumer Review Survey 2026)^[1], which is much of why the same few firms dominate every "accountant near me" answer, they are the ones actually present in the data the engine reads. Fill it out fully: exact name, address, phone and website; hours and how to get in touch; every service you offer; the sectors and client types you serve; the software you support and any partner status; and real photos with a factual description.

Genuine, recent reviews. Reviews carry weight here for the same reason credentials do, because trust is the entire basis of handing someone your finances. Recency does more of the lifting than firms expect: BrightLocal found 74% of consumers only value reviews from the last three months, and 31% will only use a business rated 4.5 stars or higher (BrightLocal, Local Consumer Review Survey 2026)^[1]. Ask every satisfied client for a genuine review at a natural moment, after a return is filed or a year-end is signed off, and make it a one-tap job with a direct link. Gather them on Google first, then the platforms that matter in your profession. Reply to every review, good or critical, but never disclose a client's financial details in a public reply: thank them, answer in general terms, and take specifics offline, because confidentiality is both a duty and a mark of a well-run firm. Never buy, incentivise or fake reviews; they are unlawful in the UK, platforms strip them out, and for a profession that trades on integrity the reputational risk is absurd.

A site the engine can actually read. Some firm websites, especially heavily-designed ones, build their text with JavaScript that only renders once a browser opens the page, and several AI crawlers never run that code, so they land on what looks to them like an empty page. It is a quiet, total failure, no journey work matters on a site the engine cannot read, and it takes two minutes to check using the self-test in our guide to whether AI can read your website.

For accountancy specifically, the readable-but-thin pattern is worth naming. QBiz Leads AI's AI Visibility Gap Report 2026, a technical readiness audit of 173 UK local service business websites, scored accountants at an all-row average of 6.85 across 46 firms, the middle of the four sectors it studied.^[4] Across the whole sample the technical basics were mostly present, but the weakest signals were the answer-oriented ones: FAQ content appeared on just 31.8% of sites and a clear process explanation on 32.9%.^[4] The audit is UK-based and descriptive, and those content figures are sample-wide rather than an accountancy-only number, but the pattern fits the three deciding factors above: most firms can be read, and the ones that get named are the ones that actually spell out the specialism, the credentials and the process in plain words.

Where the whole web has to agree about you

A researched buyer does not take your word for anything; they cross-check. They read your service page, then glance at your Google profile, your professional-body listing, your software-partner profile and a couple of review sites to see whether the story holds together. BrightLocal found the average consumer now consults around six different review sites when choosing a business, and that 97% read reviews at all (BrightLocal, Local Consumer Review Survey 2026)^[1]. The engine does the same cross-checking, faster and across more sources.

So your AI visibility is really the verdict of the whole web on your practice, plus the question of whether all of it tells one story. A firm whose name, qualifications, services, software and contact details read identically everywhere is one an engine can recommend with confidence. A firm whose ACCA listing, Xero profile and homepage each say something slightly different forces the engine to guess which version is current, and an engine that has to guess about an accountant tends to reach for one it is sure about instead. Keeping your core facts, name, address, phone number, qualifications and the services you list, exactly aligned across every listing is unglamorous and decisive in equal measure, and it is the cheapest piece of this whole job.

There is one more reason this matters more for you than for most local businesses: an accountant works to statutory deadlines, and a buyer who finds a stale fee or an out-of-date "not taking clients" line on one listing reads it as a sign of a practice that does not keep on top of detail. For your profession, sloppy listings are not just an AEO problem; they undercut the exact impression you are selling.

The free way to see which specialism questions name you

You can test all of this today without paying for a tool, and the trick is to test the filters rather than just your name.

1. Open ChatGPT, Google's AI Mode and Perplexity.
2. Put each of them the qualifying questions a real prospect would, one per axis: by sector, "accountant for contractors in [your town]"; by service, "who handles VAT and Making Tax Digital for a small business in [your area]?"; by software, "accountant who uses Xero near me"; plus the broad "best accountant in [your town]".
3. For each answer, see whether your firm is named, which firms are, and which sources the engine cited to get there.
4. Run each prompt a few times. The reply shifts between attempts, so it is the recurring pattern, not any single result, that tells you where you stand.

If your name never surfaces on a given axis, you have found the filter you are failing, the sector page you never wrote, the software directory you are not in, the credential you never made verifiable. If the same competitor is named every time on a specialism you also serve, study what they hold that you do not: fresher reviews, a clearer sector and fee page, a complete professional-body listing, a software-partner profile, details that agree across the web. The gap between the firm in the answer and your missing name is your worklist. Our AI visibility tracking guide sets out the fuller method, including the point at which a paid tool becomes worthwhile.

Where firms get filtered out

Most of the ways a practice loses are ways it fails one of the buyer's filters before it is ever properly considered:

• No Google Business Profile, or a half-built one. The likeliest reason a firm is absent from local AI answers altogether, and a common blind spot for referral-led practices that assumed they did not need one.
• Hiding cost entirely. A page that never indicates price answers none of the fee questions the engine fans out, and a buyer who finds nothing on cost reads it as a reason to look elsewhere.
• Generalist positioning. "Trusted advisers", "proactive accountants" and "bespoke solutions" tell the engine nothing it can quote and tell the buyer nothing about fit on any of their three filters.
• Thin or ageing reviews, which read as a firm that may no longer be active and fail the recency bar buyers now apply.
• A single "Services" page for everything, where a dozen offerings share one page and none of the specific qualifying questions gets a clear answer.
• Qualifications stated loosely, inconsistently, or only on your own site, when a verifiable credential is one of your two strongest assets.
• No presence in the software-partner directories, surrendering the one signal most aligned with how software-led buyers choose.
• Public review replies that disclose a client's financial details, breaching confidentiality in full view.
• Details that disagree from listing to listing, which leave the engine unsure and send it to a firm it trusts more.
• A site the crawler cannot read, because the content only loads in a browser.
• Assuming a Google ranking is enough. Ranking first on Google and getting named by ChatGPT have become two different jobs, as we explain in why a No.1 Google ranking means nothing in ChatGPT.

Frequently asked questions

How is this different from the local SEO we already pay for?

Local SEO and AEO overlap, and good SEO foundations help, but they are not the same job. SEO works to rank your website in a list; AEO works to get your firm named, cited or recommended inside the AI's answer, where there is often no list to climb. That depends as much on verifiable credentials, software-partner listings, reviews and consistent information as it does on your website, and a first-place Google ranking does not guarantee you a mention.

Do reviews really matter for a referral-led practice?

Yes, and more than referral-led firms tend to think. Even when most work arrives by word of mouth, the person who hears your name checks you online before they call, and so does the AI. Reviews are among the strongest signals both buyers and engines use, and recency is decisive: BrightLocal found 74% of consumers only value reviews from the last three months.

Should I publish my fees?

You do not have to post an exact price list, but giving a structure or a typical range answers a question almost every prospect asks and the engine actively tries to surface. A firm that says nothing about cost hands the engine nothing to quote and gives a cautious buyer a reason to move on.

Don't the big online accountancy platforms always win the AI recommendation?

Not for local, sector or software-specific questions. Location is the filter the engine almost always keeps, and a clear specialism makes you the precise answer rather than a generic one. A well-documented, well-reviewed firm with a stated niche and a verified software-partner listing can be a better answer than a national platform with no local or sector focus.

How quickly will my firm start appearing?

No set timeline applies, and the answer can change from one day to the next. The fast wins are a completed Google Business Profile and a run of recent reviews, both of which can register in weeks; the credential and software-partner listings that decide a research-led choice accrue more slowly. The durable signals are the ones that keep paying off.

Where to start

If you take only three things from this, take the ones that decide a research-led choice: make your credentials verifiable by claiming your professional-body member listing and stating your qualification identically everywhere; get into the software-partner directory for the platform you specialise in; and rewrite your key sector, service and fee pages to answer the qualifying questions in a buyer's own words. Get those right and the groundwork, your Google profile, reviews and consistent details, has something solid to support.

If you would rather have this checked for you first, a free QBiz Leads AI visibility check scans your website in about thirty seconds and returns a clear pass or fail on the key signals that decide whether AI tools can find and recommend your firm, and ranks the fixes for you. It is the cheapest place to start before you commit time or budget.

Get your AI Visibility audit →