Category: UX advice / Research articles

An evidence-based framework for redesigning your website



Are you thinking about redesigning your website? It’s somewhere on the to do list for most companies. If you’re smart you’ll know you should find some data to help you make the right design decisions—but do you know what? And do you know the most effective order to do this in?

Over the years I’ve developed a simple framework to help myself research smartly. I’m sharing it here to help you work in an evidence-based way. This way your new site design will be backed up with solid proof, not guesswork and assumptions.

By following this framework you’ll research in the right order and get the answers to the what, who, why, and how for your new design. With these answers your new site can be a meaningful improvement, not just a superficial change.

Framework for an evidence-based website redesign

The above diagram shows the core of that framework (read to the bottom to find out about a more detailed version). Start at the top (12 o’clock position) and move clockwise round the cycle, using evidence-gathering methods in the following order:

What

You should start by finding out what is going on with your website. In fact you probably already have: it’s often an poorly performing metric that motivates you to improve things. If conversions are down or you’re not getting enough sign ups then a redesign could be the answer.

If you’re redesigning because of a rebrand you’ll still want to watch the conversion data, to make sure you’ve benchmarked your site’s current performance. You don’t want a new version to make things worse.

As well as overall conversion rates and funnels, you can also gather other quantitative data on the page performance. All of this will help build a picture of the specific problem areas on your website, and where to focus your attention.

Who

Before continuing round the cycle you should take a detour to find out who your users are, especially if you don’t have a clear picture of this already. This sits at the centre of the framework because your audience should be at the heart of your work. You want to avoid designing something generic for everyone.

Dig into audience data from web analytics, survey your users to find out more about them, or outline some personas to help you understand who you are designing for. Keep them in mind throughout your redesign process—if you’re not sure whether a design is right, consider whether it is suitable for these people.

This is evidence you won’t need to gather too often, as the identity of your users *should* remain fairly consistent as you design.

Why

Moving back to the main circle of the framework and we progress from ‘what’ evidence to why. This is a core part of the process, because it’s all very well knowing what is happening and that conversion rates are down, but if you’re to have any success in redesigning you need to know why that is the case.

You can find this out from qualitative data, including things like heatmaps, visitor recordings, user testing, and analysing customer feedback. If you have a bigger budget and lots of time you can also carry out interviews or even do field research with users.

Focus on getting answers to the problems you found at the ‘what’ stage. If users are getting stuck on a page, why is that? Are they struggling to find buttons or other important functionality? Find out why and you can change things.

How

The final third of the framework concerns itself with your solutions to the issues you’ve found. At this point you should have found some pages that aren’t converting, and you learned why this is. Now you can design some solutions to this: or how you are solving the problems.

Being evidence-based doesn’t stop at the research: when you come to do the more traditionally ‘design’ bit you should also check whether your ideas are useful to users.

You can assess competitor sites for inspiration (always better to do this when you know what you’re looking to improve on your website), and you can check in on how well your designs work through lightweight design testing, prototype user tests, and perhaps A/B testing.

Auxiliary

The eagle-eyed may have spotted an outer ring on the diagram. This is to account for methods of evidence-gathering that can apply to multiple sections. Examples of these are the more esoteric like articles & blogs, client or stakeholder knowledge, expert audits, and friends & family opinions.

They sit on the perimeter around the other methods as you can use these to help with the what, why, and how at any point. However they should be used lightly as I don’t recommend them as the most objective methods.

Learn the full process

This is a high-level introduction to the order to tackle a redesign and some useful types of evidence to gather. If you’re interested in getting into the detail I’ve put together an online video course that expands on the most powerful methods of research for each section.

If you aren’t ready to hire a UX designer to help tackle your redesign, my course can help you do it yourself. It teaches what you need to know, and doesn’t require any complex software.

Sign up for a free guide to the tools you can use below or find out more details on the course here.

Last updated on 13 July 2018

ux / data / metrics / kpis / evidence / qualitative / quantitative / process / framework / course /

14 tools to gather evidence – guide

Articles on similar topics

Why you should be evidence-based in your design (and less data-driven)

8 tips for running slick remote user tests

A guide to my full evidence-based UX design process