User-generated content: how to ask for it honestly

A standard this site holds to

There are no reviews on mAutomate's own pages, and no testimonials, because we do not have any yet. Writing some would take ten minutes and it is not going to happen.

That is worth saying out loud in a post about customer content, because the advice below costs something. Search for user generated content ecommerce and you get tactics — hashtag campaigns, review widgets, incentive loops — with the awkward parts left out. The awkward parts are the subject here: what you may publish, what you must ask for first, and what you should never manufacture, however small your shop is.

Three things decide whether this is worth your time

Do you have enough delivered orders? Asking is a numbers game. Only a minority of customers respond, so with a handful of orders a month you will spend a week building a review flow to collect three sentences. Get the orders first. Getting your first customers is the prior job.

Is the product visible in use? Clothing, furniture, food, bags, prints and tools photograph well in someone's real life, and that photograph answers a question your studio shot cannot — what it looks like in a normal room, on a normal body. A USB cable does not benefit. Neither does a spreadsheet template.

Can you handle the permission and moderation work? Every published photograph needs a yes you can point to. Every review needs reading before it goes up, and a reply when it is bad. If nobody owns that, the system will quietly stop running after a month, which is worse than never having started it.

How to ask, specifically

Time it to delivery, not to payment. A review request triggered by the order confirmation asks about a product that has not arrived. Wait until it has been used: a week or two for most things, longer for anything with a break-in period, longer again for skincare where the outcome takes weeks.

Ask one question, not for a review. "Leave a review" is a blank page, and blank pages get closed. "How did the sizing work out?" or "What room did it end up in?" tells the customer exactly what to type. You also get usable sentences instead of five stars and no words, and the answers tell you what your product page is failing to say.

Remove every step you can. A link that opens a form with the product already selected, no account, no password, works. A link to a login screen does not.

Ask for the photo separately, and later. Photographs are a bigger favor than a sentence. Ask once the person has already replied to you, when there is a conversation to ask inside.

Be careful with incentives. A discount for a review biases the result, and where disclosure rules apply you have to disclose it. If you do offer something, offer it for any review regardless of what it says, never for a positive one, and say in the message that the reward does not depend on the rating. Many small shops are better off not paying at all and simply asking well.

Permission: what you actually need

This is where most shops are casually in the wrong, usually without meaning to be.

Two details people miss. First, the copyright in a photograph belongs to whoever took it, not to whoever is in it and not to whoever sold the product in it — so a customer photographing your bag owns that image. Second, permission has a scope: someone happy to appear on a product page may feel differently about a paid ad running for two years. Ask for what you need and say so.

None of this needs a lawyer for a small shop. It needs a short message, a saved screenshot of the reply, and a folder. This is not legal advice, and if you operate at scale or in a regulated category, get some.

Why you must never write them or buy them

The legal answer is that fake reviews and undisclosed incentives are directly regulated in the US, UK and EU, and the penalties are not nominal. That alone should settle it.

The commercial answer is more immediate. Fabricated reviews are recognizable — the same cadence, the same enthusiasm, the same mid-length paragraph, nobody mentioning that delivery took a week. Shoppers have read thousands of real ones and have a good ear for it. A page of clean five-star text makes a careful buyer trust you less, not more.

Then the first genuine complaint arrives, and it reads differently from everything around it. In a page of real reviews a three-star note about a stiff zip is reassuring. In a page of invented ones it is the only credible item on the screen, and the reader believes it and nothing else.

There is an operational cost too. Reviews are the cheapest product research a small store has: the sizing note that appears four times is a product change. Fabricated ones tell you nothing, and they bury the ones that would have.

And it does not scale back. A shop that invents reviews cannot later claim its reviews are genuine, because it has no way to show a reader which era they are looking at.

When this is a bad investment

Pre-launch and very low volume. There is nobody to ask. Build the page so a review section can be added, and leave it out until it has something in it.

Sensitive categories. Medical, intimate, financial. People will not attach their name and face, and asking is intrusive. Rely on specification detail, plain policies and discreet handling.

B2B and high-ticket, low-volume. One named case study, written with the customer's cooperation and approval, beats a review widget showing two entries.

Digital products. Customer photographs are meaningless. What helps is a specific outcome stated in the buyer's own words, with permission, and a visible sample.

Any store without a plan for a bad review. Not having one is the real limit. If a two-star review would panic you into deleting it, collecting reviews is not yet the right project — and deleting it is the one response that reliably makes things worse.

What to show while the reviews do not exist

The honest substitutes, in rough order of how much they do:

• Process photographs: the workshop, the cutting table, the packing bench. Hard to fake and plainly yours.
• Measurements and materials specific enough to be argued with. A page full of numbers reads as confidence.
• A returns policy in plain language, including the parts that favor you less.
• The name of the person who answers messages, and the hours they answer in.
• Your own product page quality. A page that names what the product is not does the same job a mixed review does — see the product page teardown for how that reads.

Then, as the reviews arrive, publish them unedited, including the lukewarm ones, and reply to the bad ones in public. mAutomate can draft those replies and schedule the request emails, and like everything else it drafts, nothing is sent until you have read it. The drafting is the easy part. The standard — only real content, only with permission — is the part that has to come from you.