Following on from my posts about the New vs. Returning and Frequency & Recency reports in Google Analytics, it's time to have a look at the third behavior report: Engagement. This has two parts to it – Visit Duration and Page Depth. They are both useful for a top-level analysis of how your websites users interact with the site.
This report shows how many visits lasted each length of time shown in the report. The second column shows how many pageviews each time bracket delivered.
In the 0-10 seconds row, there was a very high number of visits compared to the other rows, and the pageviews weren't much higher than the number of visits. This makes sense as those who weren't on the site very long wouldn't have had time to visit many pages. In addition, it also includes all bounce visits where only one page was viewed before the user exited.
Moving further down, you can see the visits and pageviews increased where the time frames covered a longer period of seconds. This slightly skews the view, but you can hardly expect the report to break it down into 10-second chunks from 0 to 1800+ (30 minutes) – the report would be huge and unmanageable! Grouping data is often the best way to manage it anyway, hence advanced segments.
To make this data even more useful to you than it is at first glance, make sure you utilize advanced segments. You can use these to really understand different types of visitors and what is of most value to your website:
- Visits with conversions
- Visits with conversions worth over £50
- Visits with conversions worth less than £10 but over £0.01
- Visits that included a view of a key page (i.e. a key lead generation page if you don't have monetary conversions or goals set up)
- Visits where certain files were downloaded or links clicked (measure with Event Tracking http://www.koozai.com/blog/analytics/the-complete-google-analytics-event-tracking-guide-plus-10-amazing-examples/)
From this, you might find you get a very different picture, like this example with the "Visits with Transactions" default advanced segment applied to an example profile which uses Ecommerce tracking:
Here we can see that to have completed a transaction almost all users have spent over 180 seconds (3 minutes) on the site and between them have viewed a large number of pages. If we divide pageviews by visits, we can see the average for those spending 181-600 seconds on site is a mighty 17 pages per visit.
Now, if we compare transactions with revenue over $50 to those below $50, you can see that those spending less money spend less time on site and view less pages. This makes sense because if they have spent more, chances are they had to view more pages in order to find additional products.
However, from this data you should try to understand why users are acting like this on site. See if you can encourage higher value purchases without the user having to significantly increase the length of time they are on the site and how many pages they have to trawl through.
Think about the following:
- If you make the products more accessible, could you bring down pageviews while increasing order value?
- If users have spent five minutes on your site but not converted – what's gone wrong?
Make sure, as with any analysis, that when you review this report you use it to understand your users. Ask questions of the data and your website that help you pinpoint where improvements can be made. Don't assume that every metric needs to go up – trying to increase visit duration might be counter intuitive as those on for a long time may be the ones not finding what they're looking for!
This report is all about how many pages the users viewed (i.e., how many visits saw five page views and how many pageviews did that generate in total). This report varies in usefulness depending on how many pages people view on average – when the average is high you will see a lot of data in the 20+ group that you won't be able to do a huge amount with. However, for smaller websites where the average pageviews is lower than 20, the report could be very insightful for you.
This report can be good to review based on different areas of your website. For example, blog only data may be very different to services pages. Again, it's interesting to review this for different advanced segments to compare the different levels of engagement between users who convert and those who don't.
This comparison shows results from two very different websites. See how on the Ecommerce site's graph there is a high number of people viewing more than 20 pages. However, on the blog, far less people view this many.
If you find that you have visits which triggered < 1 pageview, make an advanced segment for this. Take a look at the data in reports, such as Technology, Traffic Source and Demographics to see if they are bots or caused by something else.
Bear in mind that both reports here are based on sessions, so users coming to the site 20 times and only viewing one page each time will all fall under the 1 page row. In contrast, users who view 20 pages in one session will be in the 20 pages row.
Again, you can add advanced segments to see whether those spending more visit a certain number of pages:
This report is often proof that users view a lot of pages when placing a high value order, but perhaps your data will open your eyes to other patterns too.
For both parts of the Engagement report, the big thing is to understand what works for your users and what you can improve on. The reports can help you to profile good customer's actions, which can then help you work to get more customers of a similar nature and turn the unsuccessful visits into more successful visits.
ClickZ Live San Francisco (Aug 11-14) will bring together the industry's leading online marketing practitioners to deliver 4 days of educational sessions and training workshops. From Data-Driven Marketing to Social, Mobile, Display, Search and Email, the comprehensive agenda will help you maximize your marketing efforts and ROI. Register today!