Product pages that sell

In a shop, a good salesperson watches your face. They answer the question you have not asked yet. They know when to stop talking. On a website the product page has to do all of that alone, for a stranger, on a phone, in about eight seconds. Most product pages I audit are a photo, a price, three lines copied from the supplier and a button. That is a shelf label, not a salesperson.

The hierarchy

People scan in a fixed order, and you do not get to change it. Image first. Then price. Then the name. Then whatever is nearest the button. Then, if they are still there, the description. Build the page in that order. Put the effort where the eyes go.

Above the fold on a 390 pixel phone, that means one strong image, the name, the price with any discount shown plainly, the variant picker if there is one, and the add to cart button. Delivery time and the return rule go right under the button, one line each. Everything else can wait for a scroll.

Images

Five to eight. With fewer, the buyer cannot inspect the thing. With more, they stop looking. The first is the product alone on a clean background. That is the one that shows up in ads, search and the cart. The second is the product in use or on a body. Then detail shots: the stitching, the clasp, the label, the texture. Then scale, next to a hand or a familiar object. One lifestyle shot at most.

For clothing, size matters more than mood. A model with their height and size stated sells more than a moody editorial frame. For beauty and supplements, the ingredients panel is an image people zoom into. Make it sharp.

The description

Write it as the answer to "why this one?" Lead with what it does for the buyer, in their words. Then the material or formula. Then the fit or the dose. Then the care or the usage. Specifications go in a block that opens and closes, below, as a table, for the buyer who wants them. They are not the description.

Length depends on the price. A 600 taka phone case needs two sentences. A 12,000 taka leather bag needs the story of the leather, the lining, the hardware and the workshop. That buyer is looking for permission to spend that much.

Where reviews go

Two places. A star rating with the count sits next to the name, above the fold. It is proof at the moment of doubt. The full reviews sit below the description. Not in a tab. Not on a separate page. Photo reviews go first. A store with five real reviews with customer photos beats a store with 300 text reviews that all say "nice product".

If the product is new, say so. A "new" badge with no reviews reads as honest. An empty review block reads as unwanted.

Objections

Every product has three or four questions that stop the sale. Will it fit. Will it arrive. Can I return it. Is it real. Answer them on the page, close to the button, before they are asked. A size guide link next to the picker. Delivery time by district. The return rule in one plain sentence. For branded goods, an authenticity note with a photo of the packaging.

The FAQ block at the bottom is for the rest of the questions. Write the real ones from customer messages, not the generic ones from a template.

Considered or impulse

A considered purchase is expensive, rare and compared. Furniture, jewellery, a mattress. The page needs depth. More images, longer copy, a comparison with the alternatives. No urgency tricks. The buyer will come back three times, and a fake countdown timer insults them.

An impulse purchase is cheap, frequent and emotional. Accessories, snacks, a phone case. The page needs speed. A great first image, the price, the button, and a reason to add a second one. Bundles and free delivery thresholds do the work here. Long copy is a cost.

Most stores use one template for both. Shopify sections and WooCommerce templates make two cheap. Build two.

This week

Take your best-selling product page and your worst-selling one with similar traffic. Put them side by side on a phone. Write down the first five things you see on each. If the winner shows price, proof and delivery in those five and the loser does not, you have found the pattern. Rebuild the loser to match. Watch it for two weeks.