Short answer: a product page doesn't "sell with copy" — it removes uncertainty. People don't buy when they can't answer simple questions for themselves: will it fit, when does it arrive, what does it cost in total, what if it's wrong. Every unanswered question is a reason to leave and "think about it".
Eight questions the page answers in ten seconds
Open your own product page on a phone and honestly check how many of these are answered without scrolling and without clicking:
- Is this really the right product? — photos from several angles, in real use, zoomable
- What does it cost in total? — with tax and at least a shipping estimate
- Is it in stock? — a concrete status, not "made to order" in small print
- When does it arrive? — not "1–5 days" but "order today, get it Tuesday"
- Will it fit me? — dimensions, compatibility, specs, a size chart
- What if it doesn't? — return terms, and who pays for return shipping
- Has anyone else bought it? — reviews, ratings, customer photos
- How do I pay? — payment method icons, instalments if you offer them
In my audits the average product page answers three or four. The rest are buried in the footer, on a separate "Shipping and payment" page, or simply absent. Every gap gets closed by the customer — usually with the decision to look somewhere else.
What is missing most often
Delivery date on the product page itself. In my experience this is the single most common and most expensive gap. People choose not just a product but "will it arrive in time". An abstract "delivery in 1–5 working days" in the footer doesn't answer that. A specific date does.
Shipping cost before the cart. According to Baymard Institute, the top reason for cart abandonment is unexpected extra costs (shipping, taxes, fees): around 48%. Note what the problem is: not that shipping costs money, but that people find out at the last moment. Showing it earlier is a nearly free improvement with a direct effect on conversion.
Real photographs. One supplier render on a white background is the minimum for a catalogue and the maximum for doubt. The product in someone's hands, in a room, next to an object for scale removes more objections than a paragraph of copy.
An answer to "will it fit". For clothing that is a size chart and the model's height; for electronics, compatibility; for furniture, dimensions in centimetres rather than "medium". This is exactly where stores lose people who were already willing to pay.
Reviews. Not because "social proof" but because a review answers the questions you never thought to put in the description.
Descriptions: write for the decision, not for the search engine
The classic mistake is a description written "for SEO": three paragraphs repeating the product name plus generic words about quality. It helps neither the human nor the search engine (which has long understood the difference).
A structure that works:
- First paragraph — who it's for and what job it does. "For people who…" — this paragraph filters out the wrong buyer and reassures the right one.
- Specs as a table. People scan, they don't read. A "parameter — value" table is read in seconds.
- What's in the box. As its own block. It is a constant support question.
- Limitations, stated honestly. "Not suitable for…" builds more trust than another paragraph of praise — and it reduces returns.
Product descriptions are the place to ask your support and sales people one question: what are the three things customers ask most? The answers belong on the page, not in an email thread.
Reviews: how to get them when you have none
Stores often get stuck in "we have no reviews because we have nothing to show". What actually works:
- An email 7–14 days after delivery — not "leave a review" but one specific question: "did the size fit?"
- Asking at the moment of a repeat purchase — that person is already happy, and the response rate is far higher
- Customer photos — ask for them separately and display them separately; they convince better than text
- Don't delete the mediocre ratings. A catalogue where everything is 5.0 reads as fake. A store's reply to a three-star review sells better than ten five-star ones.
Developer note
- Product markup is not optional. Google's product structured data drives rich snippets with price, availability and rating. Stores also have a dedicated type — merchant listing — which surfaces shipping and return policy directly in search results.
aggregateRatingonly for reviews that exist. The review snippet guidelines explicitly forbid marking up a rating that isn't on the page. The penalty is losing rich results across the site.- Availability has to be true.
InStockmarkup on an empty warehouse is a straight path to manual actions and to angry customers. - The main product photo is your LCP element. Don't put
loading="lazy"on it, set explicitwidth/height, serve WebP/AVIF at the size actually needed. - Variants must not make the page jump. A colour or size switch that repaints half the page feeds straight into CLS and INP. Reserve space for price and stock status.
- The gallery shouldn't load everything at once. Eight 2 MB photos in a gallery is the usual reason a product page weighs more than the homepage.
Practical check
- Opened the page on a phone: price, stock and delivery date visible without scrolling
- Shipping cost is known before entering the cart
- At least 4 photos, including the product in real use
- Size chart / compatibility / dimensions are present
- Return terms are on the product page, not only in the footer
- Reviews exist, or a process for collecting them is running
- Product markup passes the Rich Results Test
- The main photo is not lazy-loaded and weighs under 200 KB
What's next
If the page convinced them, the visitor moves to the cart — and that is where the densest losses live. Next: cart and checkout: where 70% of ready buyers disappear.
The basis for this topic is Baymard Institute's product page usability research, cross-checked against my own store audits.