Most accounting firms we work with have between four and fifteen Google reviews and have had roughly that number for two years. The partners are not indifferent to it. They know prospects read reviews before booking a call, and they know a profile with nine reviews looks thinner than a competitor with sixty. Knowing how to get more Google reviews as an accountant is rarely the blocker. The blocker is that asking depends on someone remembering to ask, in the middle of a compliance season, when nobody has spare attention.

That is a systems problem rather than a motivation problem. Firms that accumulate reviews steadily have usually done one thing: attached the request to an event that already happens in the practice, so it fires without a human deciding to fire it.

This guide covers the moments worth attaching a request to, the wording that works for professional services, how to automate the ask through your CRM or practice software without it reading as automated, what the Competition and Markets Authority rules mean for how you collect reviews, and what to do when a review goes against you. It assumes you already have a verified Google Business Profile. If you do not, sort that first, because the review link only exists once the profile does.

Two things, and they operate on different timescales.

They shift the shortlist decision

A prospect searching for an accountant in your town sees a map panel with three firms. All three are qualified. The one with forty reviews mentioning specific outcomes reads as the safer choice than the one with six generic ones. This effect is immediate and shows up in your call bookings before it shows up anywhere else. Reviews mentioning your niche do double work here, because a construction business director reading a review from another contractor gets a signal that no amount of website copy provides.

They feed the local ranking system

Google describes local ranking as a function of relevance, distance, and prominence. Review count and rating sit inside prominence. This is slower and less controllable than the first effect, and it is worth being honest that a firm cannot review its way to the top of a competitive local map panel on its own. Reviews work alongside category selection, service listings, and citations. What reviews do reliably is stop a well-built profile from being outranked by a competitor whose only advantage is social proof.

A third effect worth noting

Language models increasingly answer questions like "who is a good accountant in Norwich" by summarising public sources. Review text is public, indexed, and written in natural language about services and outcomes. A firm with fifty reviews describing R&D claims and CIS work is legible to those systems in a way that a firm with a bare profile is not. That is an emerging effect rather than a proven channel, and worth treating as a bonus rather than a reason to act.

Timing does more work than wording. A request sent at the right moment gets a response from a client who felt something recently. A request sent at a random point in the quarter gets ignored by a client who has not thought about you in six weeks.

The moments worth building around, in rough order of conversion rate:

• A tax bill lower than the client expected. Relief plus a concrete number. This is the single strongest moment in an accounting practice and most firms let it pass in silence.
• A rebate reaching the client's bank account. Money arriving is an unambiguous positive event. Ask the day it lands, not when you file.
• An R&D or capital allowances claim settling. Larger sums and a clear sense that you did something they could not have done themselves.
• The end of onboarding. A client who has just moved from a firm that took three weeks to answer emails is at peak appreciation in week two with you. Do not wait for the first year end.
• An HMRC enquiry closing without adjustment. High stress, resolved. Handle the ask with care, but it converts.
• Filing well ahead of deadline. Weaker than the above, but usable, and it applies to almost every client at some point.

The pattern underneath all of these is a transition from uncertainty to resolution. The client felt some ambiguity about their position, and you removed it. Ask within 48 hours of that transition. After a week the feeling has faded and you are asking a favour instead of catching a mood.

What does not work: asking at invoice time, asking during January, and asking everyone on the client list on the same day in a newsletter. The last one is the most common mistake and produces almost nothing.

Professional services clients respond to requests that feel personal and specific. Four rules cover most of it.

Send it from a person

The request should come from whoever did the work, using their own name and email address. "Accounts team" and "info@" both read as bulk. If your CRM sends it, set the sender to the account manager and make the reply address go to their inbox.

Reference the specific thing you did

Compare "We'd appreciate a review of our services" with "Now the CIS refund has come through, would you mind putting a couple of lines on Google about how the process went?" The second gives the client something to write about. Most people stall on a review because they cannot think what to say, so hand them the subject.

Make the link one tap

Use the short review link from your Google Business Profile, which opens the review box directly. Do not link to your profile and ask them to find the review button. Every extra step loses a proportion of people, and on mobile it loses more than you would expect.

Keep it under sixty words

Three sentences. Thanks, the ask with the reason, the link. No paragraph explaining why reviews matter to your business, no mention of how long it takes. Long requests read as pressure.

SMS outperforms email for this in most firms we have implemented it in, because the link opens on the device the client will write the review on. Email works if that is how you normally correspond with a particular client. Match the channel to the existing relationship rather than picking one for the whole book.

A follow-up seven days later, once, is reasonable. Two follow-ups is nagging a client who pays you.

Manual requests work until the person responsible gets busy, which in an accounting practice is most of the year. Automation is what turns a good intention into a steady rate.

The build is straightforward if you already have a CRM sitting alongside your practice software.

Define the trigger events

Pick three to five events from the moments listed earlier. Each needs to correspond to something a person actually marks in a system: a job status changing to filed, an onboarding checklist completing, a refund logged. If an event only exists in someone's head, it cannot trigger anything.

Add a delay and a check

Fire the request 24 to 48 hours after the event, not instantly. An instant message the second a status changes reads as machine-generated. Add a suppression rule so no client receives a request more than once in a rolling twelve months, and an exclusion list for clients in dispute, in arrears, or mid-complaint.

Route the response

The link goes straight to Google. Do not build a pre-screening step that asks how happy the client is and only shows the review link to the satisfied ones. That practice, sometimes called review gating, is a live compliance problem under the CMA guidance on fake reviews and unfair commercial practices, and Google's own policies prohibit it. It also tends to be obvious to clients.

Track the rate, not the total

The number worth watching is requests sent divided by reviews received. Ten to twenty per cent is a normal range for professional services with a personalised, well-timed ask. If yours sits below five per cent, the timing or the wording is wrong, and sending more requests will not fix it.

Once this runs, review accumulation becomes a function of client volume rather than partner memory. A firm filing 200 returns a year with a fifteen per cent response rate adds roughly thirty reviews annually without anyone thinking about it.

Review collection sits under consumer protection law, and the rules tightened in April 2025 when the unfair commercial practices provisions of the Digital Markets, Competition and Consumers Act came into force. The Competition and Markets Authority has published guidance for businesses on fake reviews (reference CMA208) alongside a shorter guide for small traders, both available on GOV.UK.

The practical points for an accounting firm:

• Do not offer anything in exchange for a review. Fee discounts, vouchers, prize draws, and charity donations tied to a review all create a problem. Concealed incentives are specifically covered.
• Do not filter who you ask based on expected sentiment. If your automation asks only the clients you think are happy, you are publishing a distorted picture.
• Do not write reviews for clients or have staff post them. This includes a partner reviewing their own firm from a personal account, which is also against Google's policies.
• Do not ask a third party to generate reviews on your behalf. Buying reviews is the clearest breach available, and the profiles are usually easy to spot.

Alongside consumer law, your professional body's ethical code applies to how you present the firm. ICAEW and ACCA both require that marketing communications are not misleading and do not make unsubstantiated claims about the quality of your work. A review written by a client stating their own experience is fine. A review you shaped, paid for, or selectively solicited is a problem on two fronts at once.

Client confidentiality also deserves a moment of thought. If a client mentions figures or circumstances in their review, that is their disclosure to make. Your reply should not add detail they did not already put in the public domain.

Every firm that collects reviews at volume eventually gets a bad one. The instinct is to treat it as damage. In practice, a profile with thirty-eight five-star reviews and two three-star ones reads as more credible than a flawless run, and prospects read the negative ones closely to see how you respond.

Reply within a couple of days, calmly

Acknowledge the specific point raised, state what you have done or will do, and offer to continue the conversation directly. Three or four sentences. The reply is written for the next prospect reading it rather than for the reviewer, who has already made their mind up.

Never argue the facts in public

You cannot correct a client's version of events without disclosing details of their affairs, and attempting it looks worse than the original complaint. "That is not what happened" followed by a timeline of missed record submissions is the single most damaging reply an accountant can post.

Know what can be removed

Google will remove reviews that breach its policies: reviews from someone who was never a client, personal attacks, off-topic content, or spam. It will not remove a review simply because it is unfair or you disagree with it. Report through the profile and expect the process to take time and often to fail.

Dilute rather than fight

The most effective response to one bad review is fifteen good ones arriving over the following months, which pushes it down the profile and moves the average. This is another argument for the automated request system rather than sporadic asking, because it means the recovery mechanism is already running before you need it.

If the review reveals a genuine service failure, the reply is easier to write and the fix is more valuable than the review. Say what changed.

Six steps to move from occasional asking to a review process that runs on its own. Most firms can complete this in a fortnight.

Five patterns that account for most of the stalled review profiles we see when we audit a firm's local presence.