Improve your product pages

The product page is where visitors decide. They want to know if the product fits, when it arrives and what happens if it doesn't work out. Good product-page tests fill one of those gaps, or make the answer easier to find.

1. Look at all product pages as one

Varly groups similar pages automatically: all product pages count as one, and share one heatmap called Product pages (/products/…). That gives you far more data than a single product.

To check, go to Settings → Group similar pages and paste a product URL into Test a URL. It should say Counted as: Product pages. If two kinds of pages end up together, or your product pages aren't grouped, add a group under Your own groups. See Group similar pages together.

2. What to look at in the heatmaps

Open Research → Heatmaps and pick Product pages. Then work through these views:

• Attention: how long each part of the page was on screen. Where attention goes lists the parts that get the most time. Is that the photos, the description or the reviews?
• Scroll: Scroll depth shows how many visitors reach each depth, and how much of the page is visible without scrolling. Are the add-to-cart button, delivery time and review score above that line?
• First clicks (next to All clicks): what visitors click first, so what draws attention right away.
• Dead clicks: clicks on things that aren't clickable, such as photos that don't zoom, a "size guide" text or star ratings.
• Compare: Placed an order vs no order: what do visitors who order use that others miss?

Check Mobile separately: on a phone, key information often sits much lower. See Heatmaps and Filter and compare heatmaps.

3. Find the elements that sell

Open Research → Click impact, choose Product pages and sort by Biggest difference. Look at elements such as the reviews, the photo gallery, the size picker and the delivery information. A high Difference with a high Chance it's real means visitors who use that element order more often.

That's a connection, not proof. An element that sells but gets few clicks is a strong candidate to make more visible. See Click impact.

4. Watch a few product-page visits

Open Research → Recordings. Change Any page to Product pages, With and without order to No order, and With or without frustration to Dead clicks. Watch how visitors use the photos, pick a size and look for delivery information. Write what you notice in Notes. See Watch session recordings.

5. Ask what's missing

Open Research → Surveys and pick Is anything unclear about this product?. It asks:

1. "Do you have all the information you need to buy this?" ("Yes" or "Not quite")
2. Only after "Not quite": "What is missing?", with the answers "Size or fit", "Material and care", "Delivery time", "Return policy", "Reviews", "Better photos" and "Other".

It shows to 30% of visitors, after they scroll halfway down, on /product/* and /products/*. If your product URLs look different, change Show on.

To also hear why people leave without buying, add Why are you leaving?. It also shows on /product/* and /products/*.

Questions of your own work best when they are short and about one thing. For example:

• "What would you like to know before you buy this?" (Long answer)
• "How sure are you this will fit?" (Rating scale)

Keep it to one to three questions. See Create a survey and Design a survey.

6. Typical test ideas

Click Find new ideas in Ideas, and add your own with New idea. Product-page tests that often come out of this research:

• Delivery time next to the price, when "Delivery time" or "Shipping costs or time" is a top answer.
• Size or fit help next to the size picker, when "Size or fit" comes up, or the size guide has high click impact but few clicks.
• The review score near the product title, when visitors who read reviews convert better but few scroll that far.
• The return policy in one short line under the add-to-cart button, when "Return policy" is a common answer.
• A clearer add-to-cart button text or color, when first clicks go elsewhere.
• Less content above the button on mobile, when few visitors scroll down to it.

To learn which part of the page matters most, try a multivariate test. It needs more traffic than an A/B test. See Create an A/B test.

Common problems

• Every product has its own heatmap. The product URLs aren't grouped, usually because the product names are short. Add a group under Settings → Group similar pages.
• The survey gets no answers. Check that Show on matches your product URLs, and that the survey says Live. See Why is my survey not showing?.
• Click impact shows "Too few sessions". Choose 30 days or 90 days, or All devices.