How to collect reviews that convince

You've got four reviews. All four are five stars. All four say some version of "love it, thank you so much!" and one has a heart at the end.

They've sold nothing, and you can't work out why, because they're good reviews.

They're good feedback. They're bad evidence. A review only moves someone who's hesitating, and someone who's hesitating has a specific worry, not a general one. "Love it!" doesn't touch the worry. "I've had a rash from every natural deodorant I've tried and this one is the first that didn't" touches it precisely, and that one sentence will sell more $16 sticks than the other four reviews combined.

The way you get sentences like that is to stop asking for reviews and start asking a question.

What a review has to do to be worth having

A hesitating buyer is running a private list of objections. They don't tell you what's on it. For most products it's three or four things, and they're always concrete:

• Will it actually work for me specifically?
• Is it worth this much compared to the cheap one?
• Will it turn up, and will it turn up intact?
• Is this person real, or am I sending money to a stranger's account?
• What happens if it's wrong?

A review that answers one of those is worth having. A review that says "great seller, fast shipping" answers the third, weakly, and nothing else.

So before you ask anybody for anything, write down your product's three real objections. Not the marketing ones. Sit with the messages you've received and notice what people actually ask before they buy. If four people have asked "does it stain clothes?", that's an objection, and one honest review saying "three weeks in, no marks on any of my shirts" is worth more than a hundred stars.

Ask a question, not for a review

Here's the whole technique. You send a message that doesn't contain the word "review".

Quick question if you don't mind. What nearly stopped you buying this?

That's it. It works for three reasons.

It's easy to answer. People are much better at describing a doubt than at composing praise, and doubts come out in specifics because that's how they're stored.

Half the answers are product intelligence you can't get any other way. "The photos made it look bigger." "I wasn't sure the shipping was real." "I nearly bought the cheaper one on Amazon." Each of those is a fix.

And the other half have just written your best product-page line. "Honestly, the price. I'd tried three cheap ones and I assumed this was the same thing with nicer packaging. It isn't." You reply asking if you can use that, and you have a review that does actual work.

Two more questions worth having in rotation:

What were you using before, and how does this compare?

That produces comparison copy, which is what a buyer at the deciding stage is looking for and which you cannot write yourself without sounding like an advert.

If a friend asked you about it, what would you tell them?

This one reliably produces the plain-language version. People describe things to friends in the vocabulary buyers actually search with.

Ask at the right moment, and don't ask twice

The window is short. Ask a few days after they've had it, once they've used it enough to have an opinion and while it's still new enough to think about.

For a consumable, that's about a week. For clothing, after they've worn it once. For a service, within two days of the appointment. For anything with a learning curve, three weeks, because asking on day one gets you a review of your packaging.

The thing that ruins it: asking in the same message as anything else. A message that checks the delivery arrived and asks for a review reads as a pretext, and people answer pretexts with pleasantries. Send the check-in. Wait for the reply. Ask in your response to whatever they say.

Ask once. If they don't answer, that's an answer. A second nudge converts almost nobody and costs you the goodwill you'd need to sell them anything again, and how to follow up without being annoying is worth reading on where the line actually sits.

What Danielle got when she changed the question

Danielle makes natural deodorant in Columbus, Ohio. A stick is $16, shipping is a flat $4.50, and about a fifth of her orders are local pickup paid by Venmo because that's what her neighbours already use.

After 130 orders she had six reviews. All five stars, all some variant of "smells amazing, great seller!" Her conversion from her Instagram link had been flat for months.

She sent "what nearly stopped you buying this?" to forty past customers over three weeks. Nineteen replied. Eleven gave her something usable.

What came back, roughly grouped:

• Five people said the price. Two of those explained why they bought anyway, and one of those explanations became the first line under her product photo.
• Four said they'd had a reaction to a previous natural deodorant and were nervous. This was the biggest surprise, because she'd never once framed the product around irritation. She'd been selling on scent.
• Three said the shipping cost put them off and they nearly abandoned it.
• Two said they weren't sure she was a real business.

The four irritation answers changed the shop more than anything else. She put three of them, verbatim and with permission, directly under the product, with first names and cities. The one she uses at the top:

"I've reacted to every baking soda one I've tried. Three weeks on this and nothing, not even the first week. I bought two more."

Nobody had asked her for that. She got it because she asked a question that let someone describe a fear.

The other useful outcome was the shipping objection. Three of nineteen mentioning it was enough to act on, so she set free shipping over $30, and her average order went from one stick to two on about a third of orders. That wasn't a review at all. It was an interview she happened to run under the word "review".

And two people saying "I wasn't sure you were a real business" told her something no analytics tool would have. She added her city, her actual first name and a photo of her workspace to the shop, which costs nothing and is the fix for a doubt she didn't know she had. Proving you are a real seller covers the rest of that.

Where to put reviews so they do work

Reviews collected and stored are worth nothing. They have to sit next to the decision.

Under the product they're about, not on a testimonials page. Nobody visits a testimonials page. The review has to appear in the same eyeline as the price and the buy button.

Matched to the objection they answer. If your top objection is price, put the price review first. If it's fit, lead with fit. Order them deliberately rather than chronologically.

With a name and something human. "Marisa, Cleveland" beats "verified buyer". A first name and a city is enough; you don't need a surname and you shouldn't publish one without asking.

With the specifics left in. The instinct is to tidy up a review. Don't. The typo, the tangent, the slightly odd phrasing are what make it read as a real person. A polished testimonial reads as a written one.

On the message thread, too. When someone asks you a question that a past customer has already answered, paste the answer. "Someone asked me the same thing last month, here's what she said after three weeks." That's a review doing its job at the exact moment it matters, and it converts far better than the same words on a page.

Get permission before you publish anything, and be specific about what you're asking for: "Can I put that on the product page with your first name and city?" Almost everyone says yes. The ones who say no would have been furious to find it up there.

Photos are reviews too, and they're the ones you'll be short of

A photo of the thing in a real house, taken badly, is worth more than a studio shot, because the badness is the credential.

Ask for it separately and make it easy: "If you get a second, would you send me a photo of it at yours? No need to make it nice, that's the point." Roughly half won't, some will, and after four months you'll have a set of images you couldn't have produced. Using customer photos to sell covers permission and where to put them.

For anything worn or used, a customer photo also does a job your own photography can't: it shows the product on someone who isn't a model, in a room that isn't a studio, which is the single most common thing buyers say they wish they'd seen.

Incentives, and the line you shouldn't cross

You can offer something for a review. There are rules about how.

In the US, the FTC requires that a material connection be disclosed in the review itself. A free product, a discount, a payment, all count, and "I received this free" has to be visible to the reader. Buying positive reviews, or conditioning a reward on the review being positive, is against those rules and carries real penalties. The equivalent expectations exist in the UK and EU.

The practical version: never say "leave a five-star review and get 10% off". Say "leave an honest review, positive or not, and I'll send you 10% off your next one", and make sure the disclosure is on the review. Better still, don't incentivise at all at your size. Danielle got nineteen replies out of forty asks with no incentive whatsoever, because a small seller asking a direct question gets answered.

If you're tempted to write your own reviews, or to have friends write them: don't, and not only for the legal reason. Fabricated reviews read fabricated. They're too complimentary, they use your marketing words, and they all sound like one person. Buyers who spot one stop believing the real ones.

What to do while you have no reviews at all

The chicken-and-egg problem is real and it has three cheap answers.

Use messages instead. A screenshot of a customer saying "just got it, obsessed" in a DM is a review. Blur the name if you like. It's more credible than a formatted testimonial precisely because it's scruffy.

Answer the objection yourself, in public. If you don't have a review saying it doesn't irritate, write the section that explains why: no baking soda, here's what's in it instead, here's what to do if it does bother you. Self-evidence is weaker than social evidence and much better than nothing.

Give ten away to strangers. Not friends, because friends are polite and their reviews sound it. Ten people who don't know you, told plainly that you want an honest opinion and you're fine with a bad one. You'll get three or four usable sentences and at least one thing you need to fix.

And if a bad one arrives, which it will eventually, a public reply that fixes it is worth more than the review cost you. Handling a bad review goes through the wording.

One thing to know before you time the ask

You can only ask after the order is genuinely done, and on most of the rails small sellers use, nothing tells you that except you.

Sailo cannot tell you a bank transfer arrived. Only your bank can. Cash on delivery is confirmed by you when the rider hands the money over, and a Venmo or a UPI payment lands in your own notifications, not in the shop. So the "order complete" moment that triggers your review ask is a moment you mark yourself, on manual rails, by hand. There's no mobile money rail either, so if that's how most of your money moves, the shop holds the catalogue and the order and you hold the ledger.

Card payments do confirm themselves, but they sit behind the Business plan at $49 a month plus 1–3% of the goods. At 130 orders across a year that subscription is the expensive part, and it isn't the thing standing between you and better reviews.

Send the question to ten people today

Open your order list. Find ten people whose order arrived between one and four weeks ago. Send each of them, individually, in whatever thread you already have with them:

Quick question if you don't mind. What nearly stopped you buying this?

Expect four or five replies. Expect two of them to be useful. Reply to those two asking if you can use their words with their first name, put the best one directly under your product where the price is, and write down every objection you hear in a list you keep adding to.

Then do it again next month with the next ten. Six months of that and you'll have a page that answers every doubt a buyer has before they have to ask, which is the same reason getting your first ten orders is worth doing by hand: the early customers are where you find out what you're actually selling.