Optimising for eCommerce stickiness

In this post we'll be looking at a quick way to improve your eCommerce flow. In a previous life, I followed the methodology below for a global mouse-related entertainment brand and found that by implementing the slight tweaks within their purchase journey* (terms marked with * are explained in the glossary at the end of the article) , the number of journeys reduced BUT the value of each completed journey increased, resulting in a significant uptick in revenue.

The basics

First things first: you'll need to be tracking the user journeys on the site. Google Analytics is the most common solution used by sites these days. You will also need to set up your site tagging*, ideally using Tag Manager if you're working with Google Analytics. 

I would also recommend that you get an account with either usertesting.com or whatusersdo.com. Both of these services enable automated remote usability testing. Whilst they are not suitable for in-depth evaluation of a user's motivations and context, they work very well for diagnosing high-level UX issues.

The what

The first step is looking through the analytics data. Are there any particular pages where the bounce rate* is particularly high? Are there any steps within your sales or checkout process where there is a high level of abandonment? These are the types of sign you are looking for. Once you have found what is going on with your data, you are ready to find out why that keeps happening.

The why

This is where the testing platforms I mentioned earlier come in: look at where you found the issues in the analytics data. How did people get there? What could they have been trying to do? What was their goal? 

This goal now becomes the task that you set for your participants on usertesting.com or whatusersdo.com. Setting up tests on these platforms is straightforward, although we hear that the former is generally easier to use. Specify the participant details based on your target audience.

A rule of thumb is to test with at least five participants that are of the same type/ segment/persona. If budgets are limited or you are very pushed for time, we would suggest that testing with as few as three participants from your main user group should give you enough of an idea to make an educated guess as to why the issue keeps occurring.

Next steps

Your test results should now be showing you exactly what happens at that tricky point of the journey., so the next task is to try and create a solution to the problem you saw your customers experiencing on the site. For a good way to verify whether or not the proposed solution resolves the issue, you should set up an A/B test* to compare the proposed solution against the current page and verify whether your solution is better than a status quo. Google have a service that lets you do this within your Google Analytics set-up but there are several other solutions for running A/B tests on the site, including Optimizely which is favoured by a few of our clients.


Consider implementing an abandoned cart feature such as the one from MailChimp and potentially setting up a retargeting campaign* (or ten). With a cart abandonment rates* of over 70% in retail, it makes good sense to address this part of the journey.

Finally, if this all sounds fascinating but more work than you'd care to undertake yourself, get in touch with us so we can help.


A/B test - sometimes called multivariate testing, this is a way of comparing two (or more) versions of the site with a live audience to see which one achieves better results within set parameters. In this context, an improvement would be measured by the difference in revenue.

Bounce rate - the rate at which people leave the page.

Cart abandonment rate - the percentage of journeys where the potential buyer leaves the purchase journey without completing the purchase.

Purchase journey - the journey a customer makes to buy from an online store from product selection to the end of checkout.

Retargeting campaign - a marketing campaign targeted specifically to people who visited your site.

Site tagging - Tags are a type of metadata (essentially data about data) that describe a piece of information and help it to be found again. Tags in the context of this post help identify the actions that users take on the page in order to more accurately track customer journeys.