Websites have a rather confusing objective. They need to streamline the conversion process and also keep the customers on the website as long as possible.
So in short – “buy quick but hang around”.
The customer journey is becoming increasingly important but, a great first step is understanding where do people leave. This is much bigger than understanding ineffective pages which cause a high bounce rate (sessions where there is a single page viewed). So if we segment those out for the moment.
I want to look at pages which cause exits after a journey of several pages. It would also be a great idea to apply some segments then split this by the following:
- New Customers
- Repeat Customers
- Marketing Customers
- Natural Search / SEO Customers
- Direct Customers
- Referral Customers.
So now we have a list of pages which are causing exits and we also have an understanding of how those customers came to the site.
“Are those customers that come from marketing are they browsing more than purchasing? compared to Direct traffic?”
Doesn’t that sound like valuable and actionable intel to you?
What about taking the segments away and just trying to understand which pages (or type of pages e.g. category, search, product, blog, landing) are causing the highest exit rates.
Now if it’s the homepage then you could have a massive problem.
So let’s assume that you’ve done the analysis on page types. If you have a product page which is causing a large exit rate then could it:
- Be the offer
- Product Availability / Out of Stock
- Photos (are they zoomable, only offer 1)
- Video reviews (are they being viewed?)
- Is the Add to Cart button working?
Now that is something that we can get numbers on and prove to the UX designers that something needs to change.
Want additional proof?
Ever heard of session replay?
I’ve not heard of any session replay tools which look at page categories but, if you take several pages which have a high exit rate of the SAME page type look at the session replays. You can see exactly what the customers are doing.
Hopefully (and it should) complement your numbers research.