How to design a new museum website

Re-designing the Design Museum website is one of those dream/nightmare briefs, which is why we left it until the last possible moment. Our website was eight years old, our online shop probably five. The world had moved on and we hadn’t. It was virtually unsupported and after years of tinkering, tweaking and adding, it had, not surprisingly developed a nasty habit of not working. I am remembering last Christmas when the shop site went down and there was nothing we could do but count the financial losses.

This final straw really gave the project impetus and its ambitious timeline. We wanted the new shop ready for Christmas 2014. So we had a year. But we needed really to launch the shop in October 2014 to be ready for Christmas online orders.

First steps

It took us a while to get organised internally. We didn’t have the capacity to project manage the project in the existing team – and we kept going round in circles, how much and how do we configure the experience, capacity, ownership and longevity within very limited resources?  I wanted to make sure we developed something really accessible to all staff, so I was reluctant to parachute someone in, who might devise solutions that were good on paper but not workable in the reality of the museum’s set up. In the end we decided to have someone project managing at our end who had not done something of that scale but knew the organisation well and was passionate about the project. This way I hoped we would take it at our pace, learn as we went along, in our own way and end up with something that worked for us.

We adopted critical friends who had way more experience than we – thank you Jonathan Stack, Charlotte Sexton and Ben Terret. People were very generous with their time and advice – thank you George Oates, and also staff at Tate, Barbican and National Portrait Gallery who helped.

Working with an external consultant, we began by crafting a digital vision that would define and guide our ambition. Then we worked with the Directors of the museum to detail a Minimum Viable Product, anything that not included in this, but was desired, we put into subsequent phased timings. We held a ‘digi baking club’ (digital chat with cakes) with a wider group of staff to generate awareness and discussion about the project.

Setting the brief

Once we had an internal project manager in place, we started by collating the widest possible suggestions of agencies and sent them an ‘expression of interest’ form. It was light touch, designed to get insightful responses and encourage talented people to take the time to submit.  Talented people are usually busy people.

From this expression of interest form we were able to identify suitable agencies for the project and they were sent the brief – we made a decision on who would receive the brief on the basis of their scale, experience and vision. The brief document was very fulsome so to keep the project on track we included ‘the brief for the new website in three words: Engaging, Inspiring, Seamless’. We appointed a cross departmental team to sift and discuss the responses to the brief. It was very useful to get the range of perspectives and priorities represented in the decision making process, it also helped build a wide sense of ownership across the project.

We short-listed four agencies for a credentials pitch.  They came to the museum and presented on their experience, how they would approach the project and their fee structure to the cross departmental team.

As part of a larger IT review we had some recommendations about what content management system to use. And a lot of the pitching turned out to be about preferred formats and what was available in our budget range. Everyone pitching said a completely different thing. We had been burnt in the past by ending up with quite a niche technology that was difficult to support and integrate with other platforms so we were worried about having something bespoke, but at the same time something bespoke has the advantage of being something bespoke. We tied ourselves up in knots!

Kick off

After much discussion, we felt that Fabrique had a solution that while being bespoke was tried and tested.  They work in partnership with a technical agency Q42. The partnership had both the museum, technical and design expertise we were looking for. And they came across as good humoured, laid back and competent. Qualities not to be underestimated in a project of some length and intensity in my opinion.

Our project manager was thrown in at the deep end – devising with Fabrique an emergency plan in case the current website broke during the project, alongside international law contracts and setting up a series of kick off workshops in the Netherlands at short notice with all the senior staff. The workshop in Delft was the first day at the museum for our first Head of IT, a totally immersive experience for him, he couldn’t have had a better induction process.

Then it felt like weeks of timelines, timelines, timelines.


Fabrique had presented a concept at the pitch that we related to instantly: 80% of website visitors want straightforward information as quickly and easily as possible, while 20% want something inspirational or substantial.  This helped us compartmentalise audiences at a very top level and organise our thinking into either dealing with the practical needs of visitors or providing substance, which was useful for crudely organising internally. We also split ‘style’ out so we could argue about that separately, which was useful in making decisions on what something was rather than what it looked like. Oddly we didn’t disagree on style. We had a huge meeting (best attended one of the project, quite alarming for Fabrique I think: they photographed the hordes in disbelief) and we all agreed on the route for design. Thankfully, as we didn’t have the time to disagree.

Post its of ideas

Ideas for the website in post it form in Delft

The online shop was also scoped out as part of this process. We developed a criteria for selecting a shop platform.  We did some more detailed work around integration of the shop stock system, email marketing and ticketing to the sites – these work streams were run in parallel and in fact ticketing launched just after the website. We didn’t want too many contingencies on a project that was so fast paced. Any major shifts in databases that integrated to the website were parked until a forthcoming Customer Relationship Management (CRM) project. Not sure I’d do CRM and website simultaneously under any circumstances – too many unknowns, you have to have some parameters and constraints to give the project roots.

A content audit began in June that led to some photography being commissioned to fill large gaps. We ran a tree test with our closest audiences – social media, email lists and friends and family to get the broadest possible range of responses to our ideas about a stripped back navigation.

I succumbed to not having a search function on the site in late June. I had fought this, but ultimately accepted that bad search (we didn’t really have enough budget to do it well) is worse than no search. The fact that our analytics showed that on-site search was used by less than 1% of the visitors helped.

Testing, testing

On the museum’s twenty fifth birthday (5 July 2014) we tested a ‘click dummy’ with our audience panel members who have helped us think through everything from signage to programmes. We used an external agency recommended by Fabrique to get a bit of distance from the project.

We signed off final designs on 28 July. We did keep coming back to these designs as we built the site. We had strict timings and guidelines for getting content from teams. Content migration was epic. At times I had to sit for hours with colleagues just looking at pages and discussing words, pictures, links, hierarchy, meaning.

We tested everything we were given by Fabrique/Q42 in a ruthless and punctilious way – from the Content Management System to the interface. We had a fantastic volunteer who joined us for this project who helpfully had just the pedantry required. The online shop content was a struggle to get perfected within the time and resources. The capacity just wasn’t there. When we launched we had not been able to get the volume of product content we desired up, but we had to bite the bullet. And we have had the best Christmas ever online so it was undoubtedly worth all the pain.

Uploading content onto Content Management System was to the wire. The 20% element came together very late so we were all designing, sorting content, programming together in an agile fashion to ‘just in time’. We had to extend the project by two weeks which meant that our Project Manager was in Greece for the final throes. This allowed for two days of tweaking the code with programmers and reviewing content with interaction designers and learning advanced photoshop tricks from the visual designer. Something I can recommend. It really made everything 10% better. And it was good to end the project the way we started: together.

Having our pre-launch in August was great in so far as workloads were generally quieter but on the other hand, everyone on the project did have a habit of popping off for two weeks. Using an international agency was great because it forced us to be organised about physical meetings -the rest was on a google hangout for which our wifi was not up to scratch.


In the end, we let the website out quietly – but there is no quiet on the internet. It went mental. We did a naughty tweet #letscrashthissite.

Having had such a race to the finish we have spent a lot of time tidying up and we are now shaping up for the next phase of development that focusses on learning and Customer Relationship Management for the remainder of 2015. Still slightly amazed at what a small museum with a great project manager supported by the right agencies can do in a few long months!

I think if we had three bits of advice they would be

  1. Scope out and document a minimum viable product. Do this with the most senior people possible and use this ruthlessly to anchor the project as it progresses. Conversely be prepared to pivot if advised to by external expertise.
  2. Allow your design to be audience, not organisation led (this is very easy to write, difficult to really achieve, needs significant stakeholder management)
  3. A website can be won and lost in the complexity of language and content so make time to explore, think and test this (as a communications professional of course I think this) with the wider museum team
Cake and celebration at the new website's launch

Celebration at the launch of the new museum website

Leave a Reply

Your email address will not be published. Required fields are marked *