How to Write a Product Review for Something You Haven't Used (Honestly)
Most product reviews on the internet were written by someone who didn't use the product, and most of them pretend otherwise. That pretence is the problem, not the method.
A review built from the manual, the spec sheet and a thousand owner reviews is research, and research done well is worth reading; a review that invents a week of testing is a lie that readers can usually smell.
This post is the honest method: the five sources a research-based review draws on, what the page has to say about its own basis, what Google's reviews guidance and the FTC's endorsement rules actually require, the phrases that cross the line, and a structure that reads as evaluation rather than listing.
It pairs with the opinions post, which covers getting your own judgment onto the page; this is about the products you have no first-hand knowledge of.
The short answer
Build the review from five sources (the full spec sheet, the manufacturer's documentation, the pattern across owner reviews with the one- and three-stars weighted, owner forums, and a paper comparison with the alternatives).
State the basis in one sentence near the top. Evaluate, don't list: say what the spec means for the reader, what owners agree on, and who should skip it. Never use a phrase that asserts an experience you didn't have.
The five sources
| Source | What it gives you | What it can't |
|---|---|---|
| Full spec sheet (listing plus the manufacturer's page) | The facts: dimensions, materials, power, compatibility, what's in the box | Whether any of it matters in use |
| Manual and documentation | Setup steps, limitations the listing omits, maintenance, warranty terms | How annoying the limitations are |
| Owner reviews, one- and three-star first | What breaks, what disappoints, what the listing overpromises; five-stars mostly say "works great" | How common a complaint really is, without reading enough of them |
| Owner forums and subreddits | Long-term problems, the fix for the common complaint, the model owners wish they'd bought | Representativeness; forums skew toward people with problems |
| Paper comparison with two alternatives | Where this product sits: what you pay more for, what you give up | Feel, build quality, anything subjective |
Twenty to forty minutes per product, which is why these reviews take longer than a listing rewrite and less than owning the thing. The fact-check post covers checking the numbers afterwards.
Say what it's based on
One sentence, near the top, in plain words: "This review is based on the manufacturer's specifications and documentation, the pattern across about 1,400 owner reviews, and a comparison with the two closest alternatives; we haven't used the unit ourselves."
That sentence does three things: it satisfies Google's request that reviewers make clear what their evaluation is based on, it keeps you on the right side of the FTC's rule against implied endorsements, and it earns trust, because readers know most reviews don't say it.
It costs you nothing you actually had. Readers aren't choosing between your review and one by someone who owns the product; they're choosing between yours and twenty that pretend to.
What the rules require
- Google's reviews guidance asks for evaluation from the user's perspective, evidence of expertise, what the product does well and poorly, how it compares, and clarity about what the evaluation is based on. All of that is achievable from research; "first-hand" is one path to it, not the only one.
- The FTC's endorsement guides prohibit claiming an experience you didn't have. "We tested" without a test is a false claim; "owners report" is not.
- Amazon's accuracy rule for Associates says the same thing from the other side: no false or misleading claims about products.
The E-E-A-T post goes deeper on what "experience" means to Google; the short version is that honest research with a stated basis is a legitimate form of it, and faked testing is the thing that gets caught.
Evaluate, don't list
The difference between a research-based review and a rewritten listing is judgment. A listing says "2,000 lumens". A review says "2,000 lumens is more than a road commute needs and the battery pays for it: owners report 90 minutes on full, which the manual confirms."
The spec, what it means, what owners found, where the manual backs it up. That's the sentence pattern for the whole review.
Structure that forces it:
- Verdict with a reason: who it's for, who should skip it, and the one fact that decides it.
- The basis sentence.
- What it does well, with the spec and the owner pattern behind each point.
- What owners complain about, ranked by how often it comes up, with the fix if forums have one.
- Against the two alternatives, on the three specs that matter for this reader.
- Who should buy it and who shouldn't.
Phrases that cross the line
| Don't write | Write instead |
|---|---|
| We tested / in our testing / hands-on | Based on the specs and owner reports |
| After a month of use | Owners who've had it a year report |
| Out of the box it feels | Reviewers consistently describe the build as |
| I noticed / I found | The manual confirms / the one-star reviews agree on |
| Our lab / our team ran | Compared on paper with the X and Y |
Search every review for the left column before publishing. On a site that uses a writing tool, this is the single most important edit, because the tool's training data is full of the left column.
Doing it with AffSERP
AffSERP's X Best Review tool writes product sections from the five sources above (the listing's specs, the pattern across its reviews, and the comparison with the other picks), states that basis in the "how we picked" section, and doesn't produce the left-column phrases.
The owner-forum layer and your own verdict are the parts that stay with you; ten minutes per guide adds them.
FAQ
Can you write a product review without using the product?
Yes, if the review says what it's based on and doesn't claim otherwise. A review built from the specs, the manufacturer's documentation, the pattern across hundreds of owner reviews, owner forums and a comparison with the alternatives is a legitimate piece of research, and often more useful than one person's week with the item. What makes it dishonest is pretending to have used it.
Does Google penalise reviews written without testing?
Google's reviews guidance rewards evidence of experience and first-hand knowledge, and asks reviewers to make clear what their evaluation is based on. It doesn't require that every reviewer owned the product. A page that says 'based on the specs and 1,400 owner reviews' and then delivers an evaluation is following the guidance; a page that says 'we tested' and didn't is the one at risk.
How do I write Amazon reviews without buying the product?
You don't write reviews on Amazon (that's for customers who bought it); you write a review on your site that links to Amazon. For that, read the listing's full spec sheet, the Q&A section, the one-star and three-star reviews (the five-stars tell you little), the manufacturer's manual, and a forum thread or two. Then write what a careful researcher would conclude, and say that's what it is.
What phrases should a research-based review avoid?
Anything that asserts an experience you didn't have: 'we tested', 'in our testing', 'hands-on', 'after a month of use', 'I've been using this', 'out of the box'. Replace with what's true: 'owners report', 'the manual confirms', 'across the reviews we read', 'based on the specs', 'compared with the X on paper'.
Keep reading
