Answer shoppers' questions on your product pages

Check P5 in the Product data dimension, worth 6 points of 100 and judged by an AI model, which quotes its source.

What we check

An AI model reads the main content of each sampled product page, the text that agents following robots.txt can read, and tries to answer 8 questions chosen for your category. An answer counts only when the model quotes the words it relied on and the quote is found word for word on your page. The score is the share of questions answered that way.

For apparel, the questions cover fabric, fit, size range, a size chart, care, country of manufacture, the model's height and size, and the colors available. For kitchen and dining, they cover material, capacity, dishwasher safety, oven and induction use, care, weight, warranty and what's included.

Why it matters

Agents answer shoppers from your page text; when the answer is missing, they guess or recommend a store that has it.

The quote rule is strict on purpose. An answer we can't point to on your page is one an agent couldn't rely on either.

How to fix it

Add the missing facts to descriptions or a specifications block. Typical effort: 1 to 3 days.

1. Your report lists the questions your pages left unanswered. Add those facts to the product description or a specifications block.
2. Write the facts as text. Facts that only appear in images, such as a size chart drawn as a picture, can't be quoted.
3. Keep the details on the product page itself, not only on a separate page it doesn't link to.

Check it yourself

Pick a product and write down 8 questions a shopper in your category would ask. Then answer each from the page's text alone, ignoring the images.