Leaks at Consideration: The Product Page Drop
A product page leaks when a visitor who found the right product cannot assemble the confidence to add it: a missing fact, an unpriced total, an unanswered doubt. This guide reads the drop by its causes and matches each to the evidence that proves it.
The short answer
Most product page leaks are information failures, not persuasion failures. The visitor arrived interested; the page then withheld something the decision required - the total price with delivery, whether it fits their case, when it would arrive, what happens if it is wrong - and the visitor left to think, ask, or buy the clearer offer elsewhere. Diagnose by doubt, not by design: list the questions this product's buyers must answer, check the page answers each where the doubt occurs, and let recordings show you the scroll-and-stop points. The shelf bench covers the page's craft in full; this guide is the diagnostic pass over it.
Build the doubt list per product type
Five questions recur across categories, in product-specific dress. Will it fit or work with what I have (sizes, compatibility, requirements)? What does it cost in total, delivered? When does it arrive? What if it is wrong (returns, warranty, support)? And is this seller real? Write your category's versions concretely - 'fits the 2019 model', 'enough for a family of four' - and audit the page against them. A page missing one answer leaks precisely the visitors who needed that answer, which is why aggregate advice ('add more content') underperforms a doubt list: the missing answer is specific, and so is its repair.
What recordings show at this stage
Consideration leaks have visible behaviours: the scroll that reaches the delivery section and returns to the top (price assembly in progress, or failed), the size guide opened and closed three times, the tab left open while the visitor comparison-shops in others, the add button approached and abandoned for the questions tab. Each behaviour narrows the cause. Ten recordings per top product, watched with the doubt list in hand, will name most product-page leaks more reliably than any amount of aggregate data - this stage is where qualitative evidence earns its keep.
The silent pair: stock and total price
Two leaks recur above all others. Stock ambiguity: the visitor cannot tell whether their variant is available until deep in the process, and some leave at the suspicion alone. And the unassembled total: price on the page, delivery revealed at checkout, the visitor doing arithmetic they resent - or leaving to a competitor who states the whole cost early. Both are information placements, not persuasion problems, and both repair cheaply once named: availability at the variant picker, the delivered total (or its honest components) beside the price.
Comparison flight is not always leakage
Some product-page departures are the shopper's process, not your failure: considered purchases are compared across tabs and days, and the returning visitor converts at rates that flatter every page they revisit. Distinguish flight from leakage by return behaviour in your analytics (returning visitor conversion on the product) and by the recordings' tone (purposeful comparison versus frustrated searching). The repair for genuine comparison behaviour is to be the page worth returning to - the clearest facts, the saved shortlist, the honest total - not to trap the departure with interruptions.
Repair in the visitor's question order
Fix unanswered doubts in the order visitors ask them: fit and compatibility first (it gates everything), total cost and delivery next, risk answers (returns, warranty) at the point of commitment, seller proof quietly throughout. After each repair, watch the product's add rate and its recordings again; the doubt list shortens visibly. When the list is honestly exhausted and the drop persists, the cause is upstream or downstream - price positioning against competitors, or the arrival promise the page never made. The stages hand off; follow the evidence across the border.
Media doubts: when the photos are the missing answer
Some consideration leaks are photographic. The visitor's doubt is scale (how big is it in a hand, on a shelf), texture (does it look cheap up close), or fit in context (on a body, in a room) - and the gallery answers none of them, offering five angles of the same studio shot instead. The product bench's photo guide works the topic fully; the leak diagnostic is simpler: in recordings, gallery interaction followed by departure is the signature. The repairs are specific images commissioned against named doubts - a hand for scale, the product in use, the unflattering honest angle - because a doubt answered by a picture closes faster than the same answer in prose, and the gallery is where this stage's visitors ask their questions first.
Sources and evidence on the record
- Baymard Institute, baymard.com - large-sample usability research on ecommerce product pages, carts and checkout.
- Baymard Institute, research blog - published findings from its checkout and product-page usability studies.
Evidence basis: Method guide. Any figure quoted is attributed to a named published source listed below; worked examples are labelled illustrations with round numbers, not measurements. Applies when: Product pages with traffic and a weak add-to-cart rate