When should you ask for a review, and when is it too late?

Somewhere in your phone there is a customer from March who told you it was the best job anyone had done on that house. You meant to ask them for a review. Then another week happened, and then it was June, and now asking feels strange, because bringing it up means admitting you have been sitting on it for three months.

So you leave it. Multiply that by every good job you did last year and you have the real reason your review count doesn't match your reputation. It isn't that people won't leave reviews. It's that the ask kept arriving at the wrong moment, or never arriving at all.

There is a window, and it is shorter than people think

The best time to ask is while the customer can still feel the relief. Not the job — the relief. The moment where the thing that was worrying them stopped being a worry.

For most work that's the handover itself, or the day after. The job is still a story they're telling people. By the end of the second week it has become a fact about their house rather than an event, and the emotional charge that makes someone write three sentences about you has gone flat.

That's not a rule with a hard edge. But it explains why the same request gets a warm yes on Tuesday and silence a month later, from the same person, who genuinely liked you both times.

Asking straight away has its own trap

Push it too early and you get a different problem. Ask before the customer has lived with the work and you are asking them to vouch for something they haven't tested yet.

Most people feel that, even if they can't name it. They'll say yes and then wait, because they don't want to praise a boiler that might still play up. Then the moment passes and the review never comes.

So the rule is not "as soon as possible". It is: after they know it worked, before it stops mattering. For a repair, that's usually a couple of days. For something bigger — a fitted kitchen, a course of treatment, a garden — it is after the first proper use, which you can usually guess within a week.

It is not too late, but late asks need a different sentence

Here is where most advice gets it wrong. People say the ask expires, and it doesn't. What expires is the casual ask. You can absolutely go back to a customer from eight months ago. You just cannot use the same words.

A late ask has to acknowledge its own lateness, briefly, without a paragraph of apology:

Hi Sarah — hope the extension is still behaving itself. I'm finally getting round to something I should have done in March: we're trying to build up our reviews and yours would mean a lot. Two lines is plenty. Here's the link.

Three things make that work. It opens with them, not you. It names the delay in half a sentence and moves on. And it lowers the bar out loud, so they don't think they owe you a paragraph.

Expect a lower response rate than a same-week ask, and go in knowing that. Out of twenty old customers you might get five. Five is a very good afternoon.

Late asks often come back as a question instead

This is the part nobody warns you about. When you message a customer from last year, a fair number of them don't reply with a review. They reply with a question.

Actually, while you're here — do you do bathrooms as well? What would you charge to do the same on the garage? Are you still covering our village?

That is not a failed review request. That is a warm lead arriving through a door you opened by accident, and it is the strongest argument for going back through old customers even if the reviews never materialise.

It also shows something worth noticing. People carry those questions around for months without asking anyone, because asking means starting a conversation with a stranger. The answers exist only in your head, so the only way to get them is to catch you.

Writing them down changes that. A page on your site listing what you do, what you don't, where you go and roughly what things cost will quietly answer half of them before anyone has to type a message. That is the same gap hailera works on from the other side: you write the answers once, check them, and publish them as an agent your website and the AI apps people ask can read directly, so the question gets answered at eleven at night without you. The plain page is free and does most of the job, so start there and see what's left over.

A routine that stops this happening again

The fix for the March problem isn't a better memory. It's putting the ask somewhere it happens without you deciding.

Pick one existing habit and attach it. If you send a "job's complete" message, the ask goes in the follow-up two days later. If you ring to check everything's fine, the ask is the last thing you say on that call. If you invoice by email, the ask is a separate message the day after the invoice clears — separate, always, so it never reads as a second bill.

Then keep a short list. Not a system, just a note with the names of people who said yes and haven't done it. One nudge a week after, worded lightly, and then delete the name whatever happens. Chasing a third time costs you more goodwill than the review is worth.

Going back through the old ones

If you've got a year of jobs behind you and four reviews, do a clearing-out exercise once, then never need it again.

Take the last twelve months. Pick the fifteen jobs that went best — you'll remember them without looking. Write one message, personalise the first line for each, and send them over three or four days rather than all at once, so the replies don't land on top of each other while you're on site.

Then stop. The point of the exercise is to clear the backlog and start the habit, not to turn review-chasing into a second job.

Today: open your last twelve months of work and write down the three customers who were happiest. Message one of them this evening.

Written by the team at hailera — we help small businesses answer customer questions without being there.