Case study: Redesigning your website

This case study relates to the Amgueddfa Cymru website and its redesign project (undertaken between 2015-17).

In 2014 Amgueddfa Cymru – National Museum Wales realised that the historical working practises in its digital team needed to adapt to the considerable mounting pressure on digital products and outputs required. In order to manage increased requirements we would have to work smarter and rethink our old workflows. As part of this we set about planning to review and redesign the main museum website.

Jump to:

What we did | Hopefully helpful pointers

The website had expanded bit by bit over the years and had become cumbersome, difficult to navigate and update. We knew that many elements of content and design were competing with each other causing a confused offer. Our content was typically broadcast in style and many aspects mirrored the historic ‘brochureware’ (print based marketing literature) content model the site was originally set up to feature many years previously.

A screenshot of the old National Museum of Wales homepage on their website

Old website homepage featuring corporate policies and stakeholder priorities (an example of what the Museum thought was important and not necessarily what the visitors wanted to know)

Demands for publishing corporate policies and stakeholder priorities were also strong, influential and often competed for the (perceived) prime real estate of the homepage. We realised that we needed to stop, rethink and reassemble our approach if we were to retain focus on our key audience: our users.

The first and probably most important starting point is to ascertain why your website is used. What do people come to find out? We asked (via a qualaroo survey) visitors to our site why they were there and it became apparent that the majority were coming to look for information to plan a visit to one of our venues.

A table showing the primary reasons for people to access the National Museum of Wales website.

Survey results from website survey indicating most people are looking for information to plan a visit to one of our venues. Image courtesy of Graham Davies

This evidence-led approach meant we knew that no matter what other pressures we would face, our top priority must always be to make sure our visiting information was fit for purpose.

What we did

Upon establishing our priority areas (visit and venue information) we then set about to undertake a complete content audit.  We realised we needed a root and branch rethink and as our site was so complex. We systematically reviewed our information architecture to help rationalise our navigation structures and test our assumptions by thinking more like the user, as opposed to speaking like an authority.

Sounds difficult and complicated? All this means is we employed simple tactics (checked analytics and user journeys) to provide evidence for change. For instance:

  1. Printed out a physical site map showing how massive and complex the site was. This rolled up huge sheet of paper simply helped get people on board with our mission by showing them the problem physically, without necessarily having to read any of the detail.
  2. We looked at what search terms people were using when they were on the site. The most popular was the term ‘jobs’, so we highlighted current vacancies on our homepage (the number of people searching for jobs subsequently reduced, proving that we are helping people get the information they need)
  3. We looked at what devices people were using when looking at our Visiting pages (yes, websites even track when, where and what you’re using!).  We discovered that more than half were using mobile devices, so as a result, we rationalised our Visit Us information: rewriting and reconfiguring nearly 900 words down to to a suit of 11 easy to understand sentences and options providing critical information in a clear and uncluttered manner.

    A screenshot of the new 'plan your visit' page on the NMW website

    Our new ‘at a glance’ visiting page utilising icons and links to further information rather than displaying everything on one long page

  4. We discovered that the website structure had evolved as a mirror of the physical institution (department structures and collections departments). We restructured our information to be more user friendly. We drastically reduced navigation, pages and content. We rewrote content based on what people needed to know, rather than telling them what we thought was important. Critically we were given permission by Senior Management to make a call on what data could be ‘unpublished’ (no one likes the word deleted).
    A screenshot of the old National Museum of Wales website

    Image courtesy of Graham Davies

    A screenshot of the new homepage for the National Museum of Wales.

    Old homepage vs new homepage. Note the reduced corporate clutter. This has now all moved down to the footer. Image courtesy of Graham Davies

  5. We turned our homepage on its head, by simple lessening the corporate focus on our homepage (we moved corporate logos and stakeholder information to the footer) this allowed for a cleaner design upon which to showcase our venues.
  6. We didn’t need to tell people we were amazing, because they can do that for us. On checking our old meta descriptions we had the word amazing many times! We rewrote these and optimised keywords for search engines, replacing words such as ‘amazing’ and ‘fantastic’ with ‘family’, ‘Cafe’, ‘shop’ and ‘free day out’.
  7. The homepage isn’t our most important page. Again, looking at what the data from Google Analytics told us, we realised that only 3% of people visit via our homepage. Most traffic lands directly within the site depending on what search terms was used in Google (or other search engines)

Hopefully helpful pointers

  • Run a survey to establish why most of your visitors have come to your site.
  • If you know what your internal search terms are, make sure you are providing the answers to those questions people are looking for.
  • Use a website such as to make sure your content isn’t too hard to read for your key audiences. You can also check the average reading time for that content by using Google Analytics to check the average time users spend on that page. You can check if people hang around to read your content; if not, shorten it.
  • Start by building user evidence to make an obvious case for change. There will be massive pressure from others to argue that their content is the most important thing, and they won’t listen to your rationale, so you need a strong strategy and to prioritise your goals.
  • Invite your users (and staff) on your journey with you. Blog about the project, invite and involve people, give them opportunities to provide feedback and test new functionality at every stage. Also explain to people just what a beta phase is and why it matters.
  • Don’t just launch and leave. We are now constantly testing and iterating our website. After rolling out our new designs and structures we undertook fresh new user testing. Whilst a lot of the old issues had been resolved, we discovered a host of things our users found difficult. We then ran a series of updates to rectify obvious issues.
  • We will be repeating user testing regularly and adjusting our navigation, designs and workflows to match user needs.