Mendeley: Identifying issues, establishing the way ahead

Mendeley is a free reference manager and academic social network that allows its users to easily manage their research, showcase their work, connect and collaborate with over 5 million researchers worldwide.

The company serves its customers via three platforms: desktop, web and mobile (with apps for iOS and Android). But following its acquisition by Elsevier in 2013, Mendeley found itself experiencing some of the challenges that might be expected by any company transitioning from start-up to a more established business, and time to market was longer than it should have been.

How we helped

Equal Experts had already delivered standalone projects for Mendeley and had an ongoing engagement with parent company Elsevier, making us a natural choice for some exploratory work to resolve these matters. Over a six week period, we set out to identify these organisational problems in full – and to propose solutions.

Our core deliverables were to:

  • Investigate assumptions around the (perceived) status quo of the current Mendeley infrastructure and code base;
  • Enable the production of technical diagrams and documentation showing the actual current state;
  • Present recommendations that would enable the client to move forwards more efficiently and effectively.

We achieved this by conducting interviews with staff across all of Mendeley’s technology areas. The interviews enabled us to gather “brain dumps” from individuals of their knowledge of Mendeley’s technical landscape, and to get a view of all current projects in play (including their purpose, tools and processes being used, issues and successes). We also reviewed their associated code repositories.

In the course of this work we observed unnecessarily high levels of customisation and usage of niche tools; a lack of governance around whether products were fit for purpose and the ‘definition of done’; and products being driven by developers rather than customer need.

As a result, we were able to provide a consolidated list of observations and recommendations for Mendeley to move forward. At a high level, these were:

  1. Make technology more visible – The Product/Team backlogs should cover all work that that team has to deliver together.
  2. Prioritise the decommissioning of legacy systems – to remove delivery uncertainty surrounding new initiatives, projects, features and bug fixing; the time/cost of delivery is hard to define when a dependency on a legacy system is identified.
  3. Simplify and stabilise the deployment pipeline – Continuous deployment is a great goal to aim for, but it shouldn’t come at the cost of a loss in productivity, or the ability to change.
  4.  Adopt a consistent approach to similar problems – While we support team autonomy, working in total isolation is undesirable (check our blog on ‘Platform as a club’ for an example of a successful approach to this).

A clear way ahead – and a bright future

This was a slightly unusual exercise for Equal Experts, in that we did not deliver new software as part of our engagement; instead, we used our expertise in technology and process to solve a different type of business problem. Before we started, the client had an incomplete view of its overall architecture; now it has complete understanding. It all means Mendeley can now make better decisions about its technology choices in future.

And importantly for us, our client was delighted:

Beth and Dan from Equal Experts conducted a comprehensive review of our technical architecture for us at Mendeley. They adapted brilliantly to focus on the things that were most important to us, and were very good listeners when it came to interviewing members of staff. They objectively recorded the current state and made a number of sensible recommendations. Their work was high quality and we would definitely engage them again. Overall, a valuable exercise and a pleasure to work with this team.
– David Ingram, Head of Platform at Mendeley