Stall library · Evaluate
Returns or shipping policy is only in an image
A policy stall happens when your returns or shipping terms are published as a graphic rather than as page text, so an AI agent that needs to confirm the terms before recommending you finds a page with a heading and no readable words. Agents read text. A policy an agent cannot read is, for the purposes of a comparison, a policy you do not have.
Why the agent stops
Returns terms are a comparison input, not a legal footnote. A shopper asking an assistant for a jacket they can send back if it does not fit is asking a question that can only be answered from a returns page, and the assistant will answer it from whichever store published the answer as text.
Design driven policy pages are the usual cause. A carefully typeset returns graphic looks better than a wall of text and reads to a parser as an empty page with an image on it.
How often we see this
Not stated here. A frequency means nothing without the cohort it was counted over, and this page is about one stall rather than one category. Where a category has enough stores measured at every stage, the share of them showing this stall is published on its benchmark page, with the methodology beside it.
See it for yourself
- Open your returns page and try to select the policy text with the mouse. If nothing highlights, it is an image.
- Fetch the page with curl and count the words in the response. A policy page with almost no text in the source is this stall.
- Do the same for shipping, warranty and any size guide a shopper would ask about.
What to change
- Publish the policy as page text. Keep the graphic if it suits the brand, and put the words on the page as well.
- Include the specifics an agent needs to compare you: the return window in days, who pays return postage, and any exclusions.
- Do the same for shipping: rates, thresholds and delivery times as text.
- Link the policy pages from your sitemap, since they are rarely linked from navigation in a way an agent will crawl.
What Ottom looks at
Ottom fetches the returns and shipping policy pages, measures how much readable text each carries, and notes whether the page is largely an image. A short but complete policy is not flagged: a page that says the rate, the threshold and the delivery time in fifty characters has answered the question.
Questions
Does a policy published as a PDF count as readable?
Less reliably than a web page, and unpredictably so. Some agents extract text from PDFs, some fetch them and give up, and some do not request them at all. That variance is the problem: you cannot tell which kind of agent is looking at your store. If the terms matter to a purchase decision, and returns, delivery and sizing usually do, put them in HTML on a page of their own. Keep the PDF as well if customers want it to print.
How much policy detail does an AI agent need?
Enough to answer a comparison question without opening anything else. For returns that is the window, who pays return shipping, and the exclusions, which together cover most of what a shopper asks an assistant before committing. You are not writing for a lawyer or for a search engine. You are writing the two or three facts somebody would ask you in a shop, in text, where something reading the page can find them.