Two years ago I wrote a blog that explored university-business links from the perspective of a new Data Science programme at Lancaster University.

Since then our programme has enabled 70 students to experience industry placements and has led to employment for nearly all our graduates. I would like to reflect on what we’ve learnt about the factors in creating successful relationships between universities and external organisations.

Lancaster’s Data Science Masters programme was established as an industry-focussed PGT programme that would train new data scientists and prepare them for the workplace.

Ahead of the programme’s launch we consulted with Industry to establish which skills were needed in commercial data analytics and whether external organisations would partner with us to build a conduit for students to become commercial data scientists.

Our consultation revealed that companies were interested in participating, though had some reservations. The companies’ motivations differed according to their level of familiarity with data science: those new to data science wanted to conduct a feasibility study to determine what could be achieved with their data; those well-established in data science were mainly interested in possibly recruiting a new member of staff with valuable skills.

The reservations expressed entailed the readiness of students to enter the workplace and perceived complexities in working with a university. Companies agreed that universities were ideal settings for teaching technical subjects, but many were concerned that graduates lacked the ‘soft skills’ necessary to translate technical learning into useful business outcomes. Concerns were also raised about the difference in the cultures of academic and commercial worlds.

We sought to overcome the first of these misgivings by involving industry partners in the design and delivery of our programme. Companies felt that graduates lacked presentation skills, commercial awareness and an ability to take ownership of their work. After establishing the technical content of our programme, we ensured that our students had the opportunity to develop the required soft skills. On the course students are expected to organise and present their work, work on industry-specified group projects and learn how to communicate their findings to non-specialists.

Our programme emphasises that data science is not an end in itself, but is a means towards deriving better-founded answers to business or research questions. This ethos is supported by bringing in guest lecturers from industry to describe how their organisations use data science and by bringing our students to organisations where data science is making a difference (an example being developments in Wrightington, Wigan and Leigh NHS Trust to model and predict demand for A&E services).

Establishing the means for students to gain necessary skills was a partial achievement, but for our programme to succeed we needed to build partnerships with external organisations that would allow our students to use their skills off-campus. For this we needed to convince organisations of the benefits of investing both money and time in hosting placement projects. The noted differences between academic and commercial cultures primarily related to timescales (universities being perceived as less agile and responsive) and motivation (regarding intellectual property and publication).

We sought to address these concerns by establishing absolute clarity about timescales and IP. We shared a detailed process document with prospective company partners that explained exactly what could be expected from the University and when each stage of the process would happen. By receiving a clear timeline companies were able to set their expectations and get the most from the engagement.

Our motivation for running the placement programme is to provide opportunities for our students to gain workplace experience rather than to develop commercial opportunities for the University. As such, both the students and the University sign an agreement that transfers IP rights to the hosting company for anything produced during the project. Both students and companies have been satisfied with this agreement and in many cases output has been generated that is still used by the host. The only condition that the University places on access to IP is that the student be allowed to write a dissertation on their activity and submit this for marking in the normal manner.

In seeking project ideas from companies we determined that any topic would be considered valid provided it was related to any aspect of data science, was sufficiently demanding to be part of an MSc, and was related to the host’s business. We specified the last criterion, as well as asking for hosts to contribute financially for the project, to ensure that each host would be interested in deriving a tangible benefit from the project. The contribution made by each company is paid to the student as a stipend roughly equal to the minimum wage. These contributions have allowed our students to support themselves on placements away from Lancaster.

We took the decision that we would not assign students to projects, but would rather ask them to apply for the offered opportunities. This has been valuable as it gives students useful interview experience and also increases the possibility of a selected student being employed by the host after the project.

We have learnt that involving all parties in the planning and preparation of projects is essential to ensure that expectations are correctly set and that a shared view of the projects’ scope, aims and timelines is established. We ask hosts to provide the students with a line manager and we assign an academic supervisor as soon as the student is selected. We match projects, wherever possible, with the academics’ research interests as we have found that placement projects often lead to broader engagements between the University and outside organisations. Once the project team is established it works to consider whether the proposal can realistically be achieved in 12 weeks, how the most value can be derived from the project and, crucially, what data is needed for analysis. We have learnt to emphasise the last point as data access delays proved disruptive to several of our early projects and addressing this requirement has removed this issue.

Over the last two years we have learnt that the following principles are key to successful partnerships between Universities and Industry:

1. Consider what partners can gain from the engagement and how these gains can be realised. For instance, if a partner is motivated to recruit new staff offering a chance to talk to students about their organisation may be valued.

2. Consider the importance of your objectives and how compromise may be required to establish. We sacrificed IP rights as the creation of student opportunities was consider more important.

3. Establish clarity throughout – in timelines, responsibilities, and expected outcomes. We have sought to match our approach to the professional methods of outside organisations to narrow perceived cultural gaps. Setting expectations about the scope and outcomes of the engagement is vital.

4. Communicate clearly and often – in a multi-party engagement there is ample opportunity for problems to arise. When students perform placements away from the University it is important to ensure that both academic supervisors and tutors remain in contact with both student and host to ensure both progress and well-being.

We’re pleased about what we’ve achieved on our programme and are in the process of arranging placements for this year’s cohort of 39 MSc Data Science students. It promises to be another interesting year.

For more information about Lancaster University’s Data Science Programme and its placements visit:

By Dr Simon Tomlinson
Business Engagement Manager
Lancaster University Data Science Institute