"The goal of BI is to help decision-makers make more informed and better decisions to guide the business. Business intelligence software and software-as-a-service (SaaS) solutions accomplish this by making it simpler to aggregate, see, and slice-and-dice the data. In turn, this makes it easier to identify trends and issues, uncover new insights, and fine-tune operations to meet business goals."
There is no single, or simple, answer to the question ‘what do organisations want to know through their BI initiatives?’, with there seemingly likely to be as many answers as there are organisations engaged in this area. This in itself is instructive, for it demonstrates the breadth of activities which it is believed that BI can help support – thereby going someway to understanding the interest in it that it is currently enjoying. BI is clearly far from a one-trick pony.
In 2010 we conducted a survey designed to try to identify the main areas which, at the time, staff within organisations wished they had access to more information on. The results confirmed this impression of the diversity of needs amongst key decision makers within the sector, a finding also confirmed by the spread of subject areas addressed by the the eleven projects sponsored by Jisc during 2011-12.
For whilst all the projects to some degree sought to measure performance in some key area and by implication support and measure improvement, all also addressed a different need identified from within their organisations. For many of the projects the measurement was relative to a point in time or an internal target while others were comparing their organisation against external factors including benchmarking against their peer group.
Overview of project subject areas
The following summary of the main subject areas explored by the projects gives a further indication of some of the potential applications for a BI system.
The University of Manchester project sought to measure the CO2 footprint of the University and track progress in reducing CO2 emissions in compliance with UK and European targets. To achieve this they combined data from a set of existing but unconnected applications including estates management and timetable data.
The Open University and the University of Bedfordshire both sought to provide decision support tools to support efforts to reduce student attrition. In both cases the applications sought to identify changes in patterns of behaviour which could indicate that a student was becoming disengaged from their programme of study.
Both projects expect that course leaders, associate lecturers and course administrators will use this information to support meaningful and effective interventions with students deemed to be ‘at risk’ and so reduce attrition. Once a critical mass of data has been gathered it will also be possible to analyse attrition rates across programmes and identify examples of good practice which can be promulgated across the curriculum.
The Universities of Central Lancashire (UCLan), East London (UEL) and Liverpool all sought to provide senior management with strategic information of progress against key performance indicators (KPIs) by means of management dashboards. All of these projects sought to provide the critical information for senior management in a compact comprehensible format while allowing further and more detailed analysis via drill down and drill up features.
For many organisations a key question is how they are performing against other comparable organisations and, amongst others, Durham University and UEL both provided benchmarking tools and frameworks. In addition the University of Sheffield considered the practicalities and applicability of combining organisational data with open data available via such sources as the open government initiative.
For the University of Bolton the project coincided with a period of considerable change and structural reorganisation, and, as a result, the project team were subject to changing organisational priorities and conflicting demands. Within this context the project considered how best to reuse information which had previously been gathered solely for statutory returns and developed both a ‘workload allocation tool’ and a ‘costing and pricing tool’ to support staff deployment and curriculum development.
The universities of Glasgow and Huddersfield both provided BI tools to support their research effort. Both were also building on pre-existing research information management infrastructures. Huddersfield were providing tools to measure research performance against a set of parameters at a variety of organisational levels. In contrast, Glasgow were looking for methods of identifying research clusters and more specifically identifying where there were overlaps between the research interests of staff (or groups of staff) which were not intuitively obvious or derived from organisational structures.
And yet despite the disparate subject areas pursued by the projects there were also some consistent underlying themes evident. For example, many of the projects demonstrated a desire to build up a time series of data that allows meaningful comparisons and trend analysis. Likewise, all the projects met the key BI requirement of providing strategic information to inform decision making; though in common with BI projects in other sectors, for several of the projects the strategic aspect of the application was additional to a need to provide information that would improve operational processes such as student retention, meeting statutory requirements and recruitment monitoring.
Specific areas of interest
What follows is a categorised list of the main answers and areas of interest that the project organisations now have access to as a result of their BI projects. This list serves to not only illustrate the breadth of topics covered, as alluded to earlier in this section, but also to act as exemplars for organisations to consider from their own context and requirements.
- Benchmarking of performance against peer organisations at a departmental, rather than faculty level (Durham University)
- Student recruitment, application and employment by geographic area (UEL)
- Enrolment performance compared with internal targets and/or competitor organisations over specified time periods (UEL)
Student retention and progression
- Earlier and better identification of students at risk of non-completion (Open University and University of Bedfordshire)
- Identifying ‘weak points’ in modules where learners seem to be at the most risk of disengaging (Open University)
- Real time monitoring of student attendance of classroom-based learning activities (University of Bedfordshire)
- Tracking of student progression on a course or programme of study over a number of academic years, or by characteristic (eg ethnicity) (UEL)
- To connect together information about students and admissions data with external data sources relating to their place of origin (rates of deprivation, etc) (University of Sheffield)
- Access to data about the types of students being attracted to the course ie entry qualifications, previous education, location, etc, would be useful to map against performance, plagiarism and attendance and would allow academics to identify trends and future ‘at risk’ groups (UCLan)
- Accurate, open and transparent data on the time and effort spent by academic staff measured by various categories of activity (for use in Transparent Approach to Costing (TRAC) reporting and workload planning) (University of Bolton)
- Ability to identify researchers in other disciplines with whom it may be constructive to collaborate based on thematic descriptions of areas of research interest (University of Glasgow)
- Where to best target investment in research based on an enhanced understanding of research strengths (University of Huddersfield)
- What the environmental profile is of individual buildings (The University of Manchester)
- How utilities consumption and costs can be reduced (The University of Manchester)
- The impact of IT equipment in terms of energy consumption on individual buildings (The University of Manchester)
- How different timetabling scenarios affect the energy consumption and costing profiles of individual buildings (The University of Manchester)