Trust signals for accountancy firms: reviews, credentials and case studies

There is a moment, a minute or two into reading an accountancy website, where a visitor stops assessing how it looks and starts looking for proof. They have decided you might be able to do the work. Now they want evidence that other people trusted you with theirs and did not regret it.

This is a different job from the first impression, which is over in seconds and runs on design and speed. This is the evidence layer, and it is where most practices have the thinnest material. Getting accountant testimonials and the rest of your proof right is largely a question of specificity.

Reviews, and why yours are probably too old

Google reviews do double duty: they lift you in the map pack and they are the first thing a cautious person reads. Most practices have a handful, collected passively, with the most recent from two years ago.

Recency matters more than volume. Eight reviews from this year read as a practice that is currently doing good work. Thirty from 2021 read as a practice that used to. If you fix one thing on this page, make it a habit of asking, not a campaign.

Ask at the natural high point. A return filed early, a problem solved, a rebate landing, a first year completed. Send a direct link rather than asking someone to search for you, because every extra step halves the response. And ask consistently rather than in bursts, so the flow looks like a practice with steady happy clients rather than one that ran a push.

Testimonials that say something

“Excellent service, highly recommended” is not a testimonial. It is a rating with words around it, and it is interchangeable with every other one on the internet.

What persuades is specificity, and specificity usually means a small story. “We had four years of neglected bookkeeping and I was embarrassed about it. Nobody made me feel stupid, and it was sorted in six weeks.” That does real work, because it names a fear the reader might share and reports what happened.

Getting those requires asking better questions. Not “would you write us a testimonial,” which produces platitudes, but “what were you worried about before we started, and what actually happened?” People answer that honestly, and the answer is usable almost verbatim.

Attribute them properly. A full name, the business, and ideally a photograph. An anonymous testimonial is worth close to nothing, because the reader assumes you wrote it.

Case studies, in the version practices will actually finish

The word puts people off because it sounds like a project. It does not have to be. Four short paragraphs will do: the situation, what was going wrong, what you did, what changed. Half a page.

Numbers help enormously where you have them and where the client is happy for you to use them. A tax saving, hours returned, a deadline recovered. Where you cannot use figures, a before and after in plain terms still beats an adjective.

Three of these, covering the client types you actually want more of, will outperform a longer list of vaguer ones. And they are what a professional referrer reads before deciding whether to send you someone.

Credentials, stated plainly and placed properly

ACA, ACCA, AAT, ICAEW or ACCA firm registration, professional indemnity cover. Say them in plain text where people look, which is the homepage and the footer, not a compliance page three clicks deep.

The badge is not the point, the reassurance is. A visitor checking your legitimacy should find the answer without hunting, because making them hunt introduces a doubt that was not there a moment earlier.

Be careful with how you describe qualifications, though. What you can and cannot claim is governed by your professional body, and it is easier to get wrong than most people expect, which is the subject of what you can and can't claim on your website.

The signals nobody thinks of as signals

• Named people with real photographs. Professional services are bought from people. A stock image of a boardroom actively works against you.
• A real address. A practice handling sensitive financial matters that will not say where it is raises the exact question you do not want raised.
• Secure document handling. Emailing spreadsheets around is a weak look in 2026. How you collect documents is itself a trust signal, which we cover in onboarding without the email ping-pong.
• Nothing visibly out of date. A copyright notice reading 2023, or a tax figure that is no longer current, tells a visitor nobody is minding the shop. That lands badly for a profession whose entire service is being current.
• Speed. Not usually filed under trust, but a site that hesitates reads as neglected, and neglect is an odd quality to project when precision is the product.

Where to put it all

Not on a testimonials page nobody visits. Distribute it: a review or two on the homepage, a relevant testimonial on each service page, the case study nearest to the reader's situation wherever they are likely to be deciding.

Proof works when it appears next to the claim it supports. Collected in one place, it becomes a wall of praise that reads as marketing. Placed beside the specific thing a person is hesitating over, it reads as evidence.

Common questions

Can we use client names without permission?No. Ask, and ask in writing. Most clients say yes, and many are pleased to be asked. Where they would rather not be named, “a Devon-based construction firm” still carries some weight, though considerably less.

What if we get a bad review? Reply once, calmly, without disclosing anything about the engagement, and offer to take it offline. A measured reply reassures every future reader far more than the review damaged you. A defensive one does the opposite.

Is a five-star average suspicious? A perfect record across a very large number can read as managed. It is rarely worth worrying about, and it is never worth manufacturing anything. Steady real reviews are the whole answer.

Part of our guide to website design for accountants. If you have the proof but it is buried, talk to us about putting it where it actually changes minds.