Hi, a team member asked me about tracking scroll events using Amplitude.
What is the best practice on capturing events when a user scrolls to a certain part of the page?
Let’s say we wanted to have insights into whether users even scroll to the end of a page.
Do you think we should do that with Amplitude and what would be the best practice?
Best answer by dangraingerView original
Here’s a post which might give you some idea regarding page scrolling events.
Hope this helps.
This is something we do on selected areas of our main website...I say “selected” as I wouldn’t advise 1) having scroll tracking applied globally or 2) having it turned on all the time. Reason being that this can substantially inflate your event volumes for little ongoing benefit from an insight perspective...more often than not this type of insight is ad-hoc, not monitoring, so really refine your use cases first.
In terms of implementation, if you’re applying scroll tracking on a website your tag manager like has some pre-built trigger capability for “x% of page scrolled” (GTM certainly does)...so you’ll simply need to pass an Amplitude tag giving a unique name for each event (such as “50% page scrolled”, “100% page scrolled”) coupled with event properties for the page data (domain, path, etc).
The above is a quick method, something more robust (which we do) is to get your engineers to push these scroll events to your data layer and pass to Amplitude tags from there via your tag manager. You might also want to get your engineers to trigger in a slightly different fashion...things like “reached start of body copy”, “reached end of body copy” are likely more useful than arbitrary percentages (after all 50% scroll on a 5000 word page is very different to 50% scroll on a 50 word page).
If you’re applying to an app….similar advice to above for you to get your engineers to hook up.