A shopper buys a jacket rated 4.8 stars across 200 reviews. It arrives, the shoulders are too tight, and it goes straight back in the return box. The reviews were glowing. They just never answered the one question that mattered: does it run small?
That gap is where most ecommerce returns come from. Star ratings tell you people are happy on average. They do not tell the next shopper whether this specific product will fit their body, suit their skin, or work with their setup. Custom review questions close that gap. They capture the structured detail that predicts a good purchase, and the payoff lands twice: fewer returns, and higher conversion on the product pages that carry the data.
Why generic five-star reviews don't stop returns
Returns are expensive and stubbornly common. Online returns run higher than in-store returns, and apparel is the worst category for it. The reason is rarely product quality. It is mismatched expectations: the size, the fit, the color, the use case did not line up with what the buyer pictured.
A standard five-star review does almost nothing to fix that. "Love it, great quality, fast shipping" is pleasant and useless to someone deciding between a medium and a large. The review tells you the customer was satisfied. It does not transfer the one detail that would have prevented a return.
The leading cause of apparel returns is size and fit, and that is exactly the dimension a star rating flattens away. You can have a closet full of 5-star tops that all run a half size small, and nothing in the rating signals it. The information exists in customers' heads. It just never gets captured in a form the next shopper can use.
What custom review questions actually are
Custom review questions, often called review attributes, are structured questions you attach to a product or a whole category, asked at review time alongside the star rating and written comment. Instead of leaving everything to free text, you ask the reviewer something specific and store the answer as data.
A few examples:
- True to size? Runs small / true to size / runs large
- How would you rate the fit? Tight / perfect / loose
- Did it deliver the results you expected? Yes / somewhat / no
- How easy was setup? Very easy / manageable / difficult
The difference from a normal review is that the answer is structured, not buried in a paragraph. Because it is structured, you can aggregate it. A hundred reviewers answering "true to size" gives you a distribution, say 68% true to size, 24% runs small, 8% runs large. That summary is far more useful than any single comment, and it is the kind of data that drives both the buying decision and the return rate.
This is the practical core of the difference:
| Generic star review | Custom review question | |
|---|---|---|
| What it captures | Overall satisfaction | A specific, return-driving trait |
| Format | Free text + rating | Structured answer you can filter |
| Use on the product page | Social proof | Social proof + decision data (fit, results, setup) |
| Effect on returns | Minimal | Direct: sets the right expectation pre-purchase |
| Feeds structured snippets | Rarely | Yes, attribute summaries are citable |
The review attributes that reduce returns, by vertical
The right questions depend on why your customers return things. The goal is to capture the one or two attributes most tied to returns in your category, not to interrogate every buyer. Here is where the right questions matter most.
Apparel and footwear: fit, sizing, true-to-size
This is the category where attributes earn their keep fastest, because fit drives the majority of returns. The questions that matter:
- True to size (runs small / true / runs large)
- Fit at key points (shoulders, waist, length for clothing; width for shoes)
- Body context, optionally: height and usual size, so the shopper can compare against someone like them
A "runs small, order up a size" pattern surfaced from 50 reviews prevents far more returns than another paragraph praising the fabric.
Supplements and beauty: results, skin type, timeframe
Here the return trigger is "it did not work for me." Useful attributes:
- Did you see the expected results? and over what timeframe
- Skin type, hair type, or concern the buyer was solving for
- Sensitivity or reaction, if relevant
A serum that works beautifully for oily skin and breaks out dry skin needs that context attached, or every dry-skin buyer is a likely return and a likely one-star.
Electronics and tech: setup difficulty, compatibility
Returns here cluster around "too hard to set up" and "did not work with my gear." Capture:
- Setup difficulty (plug and play / some effort / needed help)
- Compatibility with the platforms, devices, or apps buyers actually use
- Technical skill level the reviewer brought to it
Furniture and home: assembly and size versus expectation
Two return reasons dominate: it was bigger or smaller than imagined, and assembly was a nightmare. Ask:
- Size versus expectation (smaller / as pictured / larger)
- Assembly difficulty and time
- Material or finish matching the photos
The throughline across every vertical: ask about the thing that, when it goes wrong, ends in a return.
How custom questions turn into better product pages
Reducing returns is half the value. The other half shows up on the product page, where attribute data does work that a wall of generic reviews cannot.
When you aggregate structured answers, you can put a fit summary near the buy button (for example, "82% say true to size"), a results breakdown for a supplement, or a setup-difficulty badge for a gadget. Shoppers get the exact reassurance they were hunting for, in the moment they were deciding, and conversion improves because doubt drops.
There is a discoverability angle too. Attribute summaries are specific, factual statements tied to a real product, which is precisely the kind of content AI search engines and rich snippets surface. A line like "this jacket runs true to size for most of its 140 reviewers" is a clean, citable fact. Generic praise is not. As more buying research starts inside AI assistants and search overviews, structured review data gives your pages something concrete to be cited for. RaveCapture collects these attributes at review time and shows the aggregated result on the product page, so the same answer that prevents a return also helps the page convert and get found in AI search.
If you want the broader playbook on using reviews to cut returns, we go deeper in this practical guide to reducing returns with customer reviews. Custom questions are the sharpest tool in that kit because they target the expectation gap directly.
How to set up custom review questions
The setup is straightforward in most modern review platforms, whether you are on Shopify, BigCommerce, WooCommerce, Magento, or another stack. The pattern looks like this:
- Find your review settings. In a typical platform you would go to Reviews → Review Form → Custom Questions (the exact path varies by tool, but it lives wherever you configure the review request).
- Add a structured question with a question type: a rating, a single-select (true to size / runs small / runs large), or a short scale.
- Scope it. Apply it to a single product, a collection, or a category. Apparel questions should not appear on a supplement, so scope by category rather than globally.
- Decide what shows publicly. Choose which attributes appear on the product page and how (a badge, a summary bar, a filter).
- Let the data accumulate, then surface the aggregate once you have enough responses to be meaningful (a couple dozen is usually enough to show a pattern).
A few do and don't lines that separate attribute data that works from data that annoys people:
- Do: ask one strong question tied to your top return reason. Don't: stack five questions and watch completion rates collapse.
- Do: use simple, mutually exclusive options (runs small / true / runs large). Don't: ask open-ended attribute questions that you cannot aggregate.
- Do: scope questions to the category they belong to. Don't: show fit questions on products where fit is irrelevant.
- Do: put the resulting summary near the buy button. Don't: bury it at the bottom of the reviews tab where no deciding shopper will see it.
If your category leans more on experience than physical fit, a short post-purchase experience survey can capture the same kind of structured signal about why people buy and where expectations slip.
Common mistakes when adding review attributes
A few traps turn a good idea into noise:
- Asking too much. Every extra question lowers completion. More questions means fewer answered reviews and weaker data. Restraint wins.
- Capturing data you never surface. If the attribute lives only in your admin dashboard and never reaches the product page, it cannot prevent a return. Collection without display is wasted effort.
- Vague answer options. "How is the fit?" with a free-text box gives you sentences you cannot aggregate. Use fixed, comparable options.
- One global question set. Fit, results, and setup are different return reasons. A single generic question across the whole catalog helps no one. Scope by category.
- Surfacing too early. Three responses do not make a pattern. Wait until you have enough answers that the summary is trustworthy, then show it.
Frequently Asked Questions
What are custom review questions? Custom review questions, sometimes called review attributes, are structured questions attached to a product or category that ask reviewers about specific traits like fit, sizing, results, or ease of setup. The answers are stored as filterable data, not just free text, so shoppers and merchants can see patterns across hundreds of reviews.
How do custom review questions reduce returns? Most returns happen because the product did not match the shopper's expectation. Fit and sizing is the leading reason apparel gets sent back. When reviews capture true-to-size feedback, fit, and use case, a shopper can set the right expectation before buying, which means fewer surprises on delivery and fewer returns.
How many custom questions should I ask per product? Ask one to three strong questions, not five. Each extra question lowers completion rates and adds noise. Pick the one or two attributes most tied to returns in your category, such as true-to-size for apparel or setup difficulty for electronics, and keep the rest optional.
Do custom review attributes help SEO and product-page conversion? Yes. Aggregated attribute data, like a true-to-size summary or a fit distribution, gives product pages unique, specific content that generic reviews lack. It also feeds structured snippets and answers the exact questions shoppers and AI search engines ask, which supports both conversion and discoverability.
Which products benefit most from custom review questions? Any product where the return reason is predictable and expectation-driven: apparel and footwear (fit, sizing), supplements and beauty (results, skin or hair type), electronics (setup, compatibility), and furniture (assembly, size versus expectation). The more a return depends on fit or expectation, the more attribute data pays off.
The takeaway
Returns are mostly an expectation problem, and generic five-star reviews do nothing to solve it. Custom review questions capture the one or two structured details that predict a happy purchase in your category, then put that data where it does double duty: setting honest expectations so fewer orders come back, and giving product pages specific, citable proof that lifts conversion.
Start small. Pick your single biggest return reason, write one strong structured question to capture it, scope it to the right category, and surface the aggregate near the buy button. One good question, shown in the right place, will move returns more than another hundred reviews that all say "love it."
