Should I care about average engagement time?

Google Analytics throws lots of metrics your way. If you're looking at how engaged your users are, it's easy to latch on to average engagement time per active user. It's the first number in the Engagement Report, after all.

Is it a good metric to track? Or is it prone to distortion?

Imagine you have a content website - like BBC News - and you provide stories for your audience. If your content is good they will read more, would you agree? And if it's boring, or confusing, or broken, they will spend less time there, right? 

That's a good general rule. 

Let me give a caveat. Imagine your aim is for someone to understand your message. That might be an aim for a public service website, like the NHS portal in the UK, or a local council website such as Liverpool City Council. A long-winded, badly-worded page would hinder that goal. Confusingly, a long-winded, badly-worded page might also have a longer average engagement time in Google Analytics (GA). 

Imagine you work for a cancer charity and you're trying to raise money. You run an email marketing campaign. The emails lead supporters through to a particular website page to give money. But imagine that the donation form is actually stored on a different website such as the Charities Aid Foundation.

In this context success is measured in the number and quantity of donations. That money helps the charity invest in cancer research, and provide support to those affected.

These would be the steps in the user journey:

1. Email
2. Information page on website
3. Donation page on CAF Bank website
4. Receipt and thank you page on CAF Bank website
5. Thank you email

Now, here's the contradiction. If more people make a donation, more people will leave the charity's website. And that will bring the average engagement time down in GA. Darn.

I've seen an extreme example of a website like this where the average engagement time was 36 seconds. The owners weren't worried about that - they had other data that their users were achieving their aims, it's just the end point of those aims was a separate website.

Cross-domain tracking can help in that sort of situation. Sadly, in the example above it wasn't possible: the owners didn't have access to the back-end of the fulfilment website.

My advice is to think about the purpose of your website or app. Think about the kind of pages that form it. And then decide whether the average engagement time metric fits for you.

Also, remember that average engagement time is a site-wide metric. As such, it may take a huge amount of effort to shift. Perhaps it would make more sense to focus on average engagement time for a single page instead. You can find that in Reports -> Engagement -> Pages and screens.

About me

I'm James, a Google Analytics consultant in the UK.

I can help you make the most of Google Analytics.

Learn more