Category: UX advice / Research articles

How to watch and analyse session recordings to gain powerful insights


“Wait, those are real people who were on the site just now?”

“Yes.”

“And you can see everything they do?”

“Yep, spooky hey?”

“Er yeah! How does it work?”

There’s still a bit of magic that comes with first seeing playbacks of website visitor recordings. I used to have conversations like the above a lot when demo-ing them in workshops. Still do occasionally, but people have become ever more aware of them in recent years.

There’s an increasing number of tools that offer the ability to see the exact journeys and clicks that your users make (HotJar, Lucky Orange, Inspectlet, FullStory, etc). More and more websites that I work with have such a piece of software installed, gathering reams of insightful data.

The challenge is knowing what to do with these recordings.

Get watching

There’s no getting away with it, to get anything useful out of these recordings you’re going to need to actually watch them. Preferably a bunch of them in one sitting. It’s by doing this that you’ll start to spot insights.

It’s a good idea to filter through them to look at only a few with certain characteristics (especially if you’ve got a lot to get through). For example, if you’re working on redesigning a certain page, just watch all the sessions that include that URL and see what the common factors are.

When looking through the videos, the aim is to build a picture of the common user behaviours that you witness. When you see something interesting happen make a note of it and then tally up each time you see that again in future. I find watching 50 videos is a good amount to get a sense of repeated patterns.

But just what should you be looking out for?

1. The order of actions

One of the benefits of visitor recordings is that you’re getting the dimension of time being played out in your data. If you’re just tracking events you can see whether or not people use different elements like filters, but with recordings you can see the order they use them in.

This is helpful as you can see if they’re using them in the order you designed or assumed. Are people actually doing things completely in reverse? Perhaps you have two elements on a page that are doing a similar job and could be combined to make it easier for users to know the right thing to choose.

2. Precise engagement

It’s good to know that your image galleries engage your users more than blocks of text, but with visitor recordings you can see exactly what images are turning people on. You get to learn the specifics of the content that is working, rather than generically knowing that most body text is skipped over. It’s likely that some is working better than others.

3. Dwell time

Quantitative analytics can tell you engagement metrics like average time on page but that doesn’t give the full picture. Your analytics could show that Page A has an average time on page of 3 minutes, and page B has 2 minutes. Stopping at this high level makes you think that A has 50% higher time on page.

However if you watch what people are actually doing then you might see that users spend 3 minutes browsing on Page A before going to Page B for 2 minutes, and then returning to Page A for another 3 minutes. Now you can see that in a journey they’ve spent 3x as long on Page A as B, showing that this is far more popular with users.

4. The journey

Users journeys are all important in UX design. It’s how we build up the full experience and design websites. We usually construct our conversion funnels to reflect that, so we can track how many people reach step 1, step 2, step 3, etc.

What’s harder to find out is those alternative journeys that people are making, because in a lot of software as long as they hit all the pages they’ll be recorded as having done the journey. However in reality we don’t know if they’ve gone from 1 to 2 to 3 to 4, or if they’ve gone 3 to 1 to 2 to 4.

We also don’t quite know if they’re journey includes other pages on our site that we’re not tracking in the main funnel. They could have gone from 1 to a page like FAQ back to 1 then to 2 then to another page like delivery information, before 3 and 4. This is the fuller picture we get from absorbing the videos and helps us know whether content on those extra pages should be incorporated into the main flow.

5. What’s ignored?

In your rush to record everything that users do, you can also forget to take notice of what they don’t do, which can be one of the main benefits of visitor recordings. By spotting what is routinely ignored by users you can find candidates for features to remove from your site altogether.

Removing clearly unimportant stuff should only make your pages more focussed and your website easier to use.

6. Missed messages

On the subject of watching what users ignore, a more specific thing you can find out with visitor recordings is whether they are seeing messages you put up to users. Of particular use is to know whether people are spotting error messages when filling out forms.

Users will always make mistakes on forms, the question is whether you have built them robustly enough so that they can then correct those mistakes themselves. I’ve watched plenty of recordings where the error messages get missed and the user ends up completely stuck. Good error messaging can feel like an unexciting thing to design but it’s often the thing that determines whether users stay or leave your site in frustration.

Watch and learn

If you do nothing else, set aside a few hours to play a set of these videos and absorb the events you see. Internalising them will help you know what’s working and what isn’t. It may seem like something that will take ages but as you get more attuned to watching them on a certain website you can up the playback speed from 1x to 2x to 4x. The unusual will always jump out.

Of course, recording the events and having a more precise tally will help you have evidence to show others and refer to in future—something I thoroughly recommend. But whatever you do, don’t ignore any recordings you have. There’s gold in there.

Last updated on 6 November 2019

Articles on similar topics

A guide to my full evidence-based UX design process

Copying, stealing, and inspiration: how to do competitor research properly

Small to medium sized websites shouldn’t bother with A/B testing