A hotspot sits over the delivery note on today’s product page. Before the last update, the size selector occupied that position. If the map combines both periods, the coordinates alone cannot tell you which element visitors clicked.
In a position-based heatmap, the click data must match the displayed page version. Before redesigning the delivery note, establish whether the clicks shown were recorded against that layout at all.
One address can contain several views
Fictional case: On Wednesday, a clothing shop adds a promotion above the product name. The size selector moves down. On Friday, someone opens the whole week’s heatmap over the current page. Older clicks now appear over different content.
The page can also change without a release. An open size guide, a late-loading image or a sticky purchase bar changes how positions relate to elements. A shared URL does not establish that the page was in the same state.
PostHog’s documentation explains that allowing a wider range of viewport sizes can reduce positional accuracy and that fixed elements need separate handling. This describes PostHog; it does not establish equivalent functionality in another tool.
Narrow down the area in question
Compare the release date with the selected date range. Separate data from before and after the change where your collection allows it. Then compare viewport width and visible page state. “Mobile” is often too broad: a wrapped title changes the height even within one device category.
| Situation | What you need before interpreting it |
|---|---|
| New promotion above the product | Matching date range or documented layout version |
| Open dialog | Evidence of whether the points belong to the dialog or the page behind it |
| Sticky purchase bar | Appropriate handling of scrolling and the fixed element |
| Several product pages in one map | Comparable structure and the same meaning for the target area |
If the required version is missing, the old mapping may be impossible to recover. Mark the area as unresolved rather than moving points onto today’s button by eye. A documented layout identifier can help with future collection, provided the tool actually supports it. Customer names and form entries are unnecessary for that purpose.
Turn the map into a task someone can check
A useful finding reads: “Cluster near the former size selector; the current delivery note is not established as the target.” Examine the selection journey on the matching version or with a purpose-built example. Once the target is identified, you can ask whether people find the help they expect there.
If the displacement happens only while the page loads, investigate layout shifts next. The Purchase barrier check addresses the clarity of a product page and its basket. A heatmap by itself establishes neither a cause of abandonment nor lost revenue.
The scope of SAND & MEHR’s tool is described under Analytics & A/B Tests. Check the limits of the available views there before planning an analysis for your website.