I measure Instant Articles with Google Analytics for quiet a while now and doubt the results of some of the metrics recorded.

For instant articles the average time on page is 00:03:30 where normal mobile traffic from facebook is only about 00:01:00. Given that instant articles are short and simple that strikes me as to high. Also bounce rate is about 90% which leaves a remainder of 10% to calculate the average time on page; rather unreliable.

I also queried facebooks aggregated Instant Articles API and the average time is very different with about 00:01:20.

Did you experience similar results? Do you have any explanation for the stated data or an alternative idea on how to measure performance reliably and comparsion with regular mobile Facebook traffic in mind?

PS: All information are backed by data from round about half a year.

  • Have you implemented Google Analytics events for scrolling yet? Sending GA data when the user scrolls down the page to read the article is a great way to track how long users actually spend on your pages. – Stephen Ostermiller Dec 9 '16 at 11:50
  • I didn't, but I will in the near future. I'd still be happy to understand the discrepancy beteween data collected by GA and FB Insights. Built in time based metrics seem unreliable in genera when it comes to GA. – SomewhereDave Dec 11 '16 at 11:06
  • By default GA measures time between the first page load and the start of the last page load. One page sessions have zero time. The only way to improve it is to implement events that send GA information as users use your pages. – Stephen Ostermiller Dec 11 '16 at 11:12
  • I'm aware of the calculations as stated in Google's documentation, I'm still in doubt about the actual numbers though. That's why I'm looking for a way to assess the numbers I see. If there was a way to debug the tracking itself, that would be very helpful. – SomewhereDave Dec 19 '16 at 22:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.