Analytics & Conversion Optimisation

< Back to CX/UX Research

Design without research is largely guess work. This is where analytics come in. Tools such as Google Analytics provide a mountain of information about what is happening on your site, and allows you to truly understand what's going on.

Analytics tools can tell you:

  • Where you are losing traffic
  • Where users are getting stuck, or heading in the wrong direction
  • Which pages users are looking for, but not finding (problems in the Information Architecture)
  • What content users want, that you may not even offer (yet)
  • Where the key funnels are, in terms of user behaviour
  • Where they are broken
  • Where they are leaking
  • How much money those leaks are costing you.

We can take those findings and work with your team to define solutions and pick out the gold. As with all our services, it really depends on what you need from us and from your users.

Conversion optimisation – finding the putty

Conversion optimisation is about finding the tweaks and design solutions that slowly plug those leaks. Taking each step of the funnel one by one we analyse activity and behaviour, and apply treatments that are likely to have a beneficial effect. That can be measured, and the process is simple; if conversion goes up, we keep the change and move on to find more holes. If not, we simply rewind and look for more solutions.

Multi-Variate Testing (MVT)

Multi-Variate Testing, or MVT is an excellent tool that allows you to measure combinations of design changes in a single test. We identify a range of elements that can change on a page, and all these changes are applied in a matrix of variable design changes. For example we may identify that the page might benefit from:

  • Reduced text description – from light to medium, low, or non-existent
  • Larger Call to Action buttons – we might try larger, and really large
  • Removal of one or two optional fields
    • A Multi-Variate Test creates pages that have differing combinations of all of those changes, and then serves them to your customers on a random cycling basis. It measures conversion performance for each version, and then at the end of the day (or longer, if you require) it reports on the winner. For example it may be that the page with the least explanatory text, the largest button and the removal of both optional fields has the highest conversion rate of all.

That then becomes the optimised version you keep. Simple.

A/B Testing

This is a simpler approach whereby two versions of the page are created – usually the existing design, and an improved design. Both are randomly served to users on a cycling basis, and their conversion is tracked. If the improved design has higher conversion, it’s on the right track.

Even if you have no specific optimisation tools available, we can easily help you to improve your conversion – as long as you are tracking it. And if you aren’t tracking it, we can help you set that up, too.

< Back to CX/UX Research