Web Retreat Number 3 – EMBL Communications

EMBL Communications

Informing, inspiring, and engaging society with EMBL’s research, services and training

Web Retreat Number 3

Last week, the Digital team – together with a few people from Heidelberg IT – spent a couple of sunny days at EBI with our friends in Web Dev. The aim of this third retreat was to discuss and plan the EMBL.org roadmap and identify projects that would address three core areas: technical infrastructure, web development, and enterprise data. Let’s take a close look at each of these areas.

Technical infrastructure

As a team, we’ve decided on using EBI’s existing supported technical infrastructure for the new EMBL.org. This means that, in 2019, when Fiona – our current content management system – is retired, all of our websites will be hosted from there.

One of the outcomes of the last retreat was that we will be using WordPress as our go-to solution for small websites and blogs. This is already a supported solution at EBI. During this retreat, we dug into how our WordPress solution works technically, together with how we might use WordPress for our many needs during the next two years.

Web Development

The digital comms team is currently a couple of months into redesigning and developing the new Visual Framework (v2.0). This new system comes from expanding the scope of EBI’s current Visual Framework and is being headed up by Ken, supported by a freelance front end developer, Stu. Together, they have made great progress in defining the core architecture of the system and how we might scale it for the whole of EMBL.

During the retreat, the team was given an update on the system and the current roadmap.

Enterprise data

During the last retreat, we established that the IT team in Heidelberg would be responsible for enterprise data. This is organisational data that represents, for example, people, jobs, publications, taxonomy etc. A goal for this retreat was to figure out where to begin in understanding the differences between EBI and EMBL enterprise data and building a solution that addresses those differences. Once we have that, we can build a web service that could provide that data to web sites and applications.

A new solution for Group Pages

After working through those three topics for best part of a day, it became clear that a project that would tackle all of those would be the Group pages on the website. Groups include content on people, jobs, publications, services, and editorial content about the group.

The Group Sites project can be summarised like this.

Our Minimal Viable Product (MVP):

  • Each group in EMBL will have an automatically generated WordPress installation as part of the new EMBL.org.
  • This WordPress site will be populated automatically with information from the existing official channels (for example, EMBL.de).
  • The automatic content will come directly from enterprise data sources: People, Jobs, Publications, Intro content.
  • Crucially, for Groups, this means the content is *the same as today* but with the addition of group-specific jobs. All of this data is automatic, which means there is no administrative overhead for groups.
  • As this is built on-top of EBI’s existing infrastructure, it has the benefits of scale. That means security updates, hosting and backups, and a team people to field technical support issues.

In the near future:

But we do know that many groups need more than Fiona can currently provide. So, they make their own websites. This is fine. But it means that content is created and maintained in a vacuum. Wouldn’t it be great if groups could share content resources? Or be able to publish directly to EMBL.org? The new WordPress-based platform will allow this and more. Here’s an overview of where this could go:

  • Groups will have the option to ‘adopt’ their website. When they do this, a primary user will be assigned administrative rights and be able to add and modify content beyond the automatically generated content.
  • We understand that many groups and teams have their own design requirements. This could be as minimal as their own logo, or a colour change. Importantly, this is an opportunity for groups to feel they have a degree of control of how their content and group is presented to users.
  • Some of the features that were discussed were: Pages to show software; collaborations; news (which could be syndicated across EMBL); Trainee/Internships; Alumni (with integration with ThankQ); Team blog or updates; Social widgets.

Those groups or teams with existing external microsites can choose to move to this infrastructure. Or they can remain on existing solutions. If they choose to do that, however, it will be without EBI or Stratcomm support.

Tackling this project also helps with the new Visual Framework. By working alongside the development of the WordPress theme, Ken and Stu can work with our WordPress theme provider to create new content patterns. In turn, Ken and Stu can help inform the coding decisions made in WordPress. This cyclical process – where two streams of work inform one-another – can not only dramatically increase the the delivery of a project, but can also result in better, peer reviewed code.


  • We need a new WordPress theme for Teams and Groups
  • We already have a web service for Jobs and Publications – Managed by the team in Heidelberg IT. But we need rationalised enterprise data for People. This new enterprise data service will rationalise the differences between EBI and EMBL and present a single source of truth. We’ll then use this to build the People pages automatically.
  • We need to communicate and consult around this change. From a data perspective, actually, not much changes with the exception of adding Jobs which are specific to a given group. It will be the same content currently in Fiona. But this existing data and content will be presented within the new design and context of EMBL.org
  • A process for consultation. Not only do we need to consult with groups around this change, we need to figure out how to do it at scale. With ~200 groups and teams, maybe a thousand people, and across 6 geographic locations, it’s a big task for a tiny delivery team. And at that scale, it’s my experience that conversations don’t cut it. People don’t scale. We’re a small team and we need to work on a process by which we can gather useful feedback from our internal stakeholders and users at the same time as being able to still produce the tools and deliver. Quite a job.

What about the Landing Pages?

Oh, nothing changes there. We’re still working on them, and still planning launch by end of this year. The great thing about the landing page project is that we have many of the building blocks to produce the WordPress theme from the work Stu and Ken have done.

The Content Hub will start to house some of the content. The Visual Framework will crate the various building blocks of the new WordPress theme. And EBI’s infrastructure will host them.

The messy middle

Many website projects I’ve worked on over the years follow a similar trajectory: organised planning, project discovery and research leading, eventually, to delivery. But there is a bit in the middle. A bit that often takes people by surprise. I call this the messy middle. It’s a stressful time when a project team can be seen to making more of a mess than they are fixing. A time when stakeholders can feel the project is aimless, or not delivering, or isn’t where it should be. It can be a time of panic and stress as deadlines approach. But, for me, it’s where the real work happens.

We started the EMBL.org project about nine months ago when I joined EMBL. I think we’ve had a pretty good run of it so far. We’re in that first bit, still. Making quiet progress. But things are about to change as we move into the messy middle.

The projects we’ve tackled this year – The landing pages, Group Pages, taxonomy, Content Hub, Visual Framework, content audit, Springboard – are all pieces of one big machine; the new EMBL.org. We’ve been spending our time addressing thorny infrastructure issues. Some of these are big, complex issues with their own history and baggage. We couldn’t just start and ‘build a new website’, or even hire a company to do that for us. I knew this would be the case, EMBL.org is not a website project. They never are. This is a project about transformation and it goes to the very heart of the organisation. It is inherently messy. From its data to its people, digital transformation has to be tackled holistically and concurrently if it stands any chance of success. That’s what we’re going to carry on doing.

Until next time

The next web retreat is planned in November or December of this year. I’ll follow up with an update then.