Case Study - Varied Vendor Approach
Unsatisfactory systems
BI Maturity Model Stage 1
New senior management in a College or University [The Institution] found that their systems for managing financial and student data were not satisfactory. Financial data were not entered correctly; were not available when needed; and did not agree with spreadsheets kept by individual schools or departments. Student data were inaccurate; student names were duplicated, misspelled, and omitted. Central student records often disagreed with student records kept by schools or departments. Meetings of senior staff in the Institution often spent half or more of their time discussing errors in data and working out which data to accept.
A set of individual systems
BI Maturity Model Stage 2
The new management of the Institution set about replacing their systems with new, fit-for-purpose systems. The Finance systems were replaced first. The new system was selected from a vendor who had experience of the Education sector, and for whom good reference sites were available.
The Curriculum information, HR and Payroll systems were replaced next. Again, systems were selected from vendors with good reference sites in the Education sector.
Finally the Student information system was replaced. A different vendor from those above, but with a good track record in Education was selected. As with the earlier systems, the Institution's senior management selected vendors with whom they could establish a good working relationship. The initial 'vibe' was important, and this was checked by making sure that the vendor sent the same technical people to demonstrate the system as would be working with the Institution to implement the system.
This work led, over a period of several years, to a set of centrally trusted, fit-for-purpose systems.
Tips: system selection
Visit users and see good system demos, related to your own situation
Examine more than one architecture
Think about your long-term relationship with the vendor (even if it is an internal vendor)
Set a realistic schedule
Devolved, Supported Responsibility
BI Maturity Model Stage 2
At the same time as new systems were being introduced centrally, local administrators and Deans, Heads of School, or Heads of Departments were trained to use the new systems. This support was clearly linked to local responsibility for correct, timely, accurate and complete entry of data. It was also linked to local access to the central data systems, and therefore to the end of locally held spreadsheets and databases.
Now all staff were trained to use the central systems, and were expected to use the central systems. All data were entered once, and once only, into the central systems. All staff whose duties required it, had access to the data in the central systems.
Data validation routines were put in place where possible. These were either format masks (so a date field had to have a correct date; a cost field had to have a correct cost of the right size) or drop down lists.
Since the local managers and administrators were clearly responsible for the data quality from their unit, there was no longer any scope for debating data quality at management or planning meetings. Any criticism of data quality was a criticism of the manager whose unit had entered the data. If the data were indeed faulty, that manager had to repair them and improve their processes. Suddenly meetings spent only a few minutes discussing data quality issues. Almost all of the time was spent on the real business of each meeting. This alone represented a big increase in efficiency and productivity.
Business Intelligence (BI) System Selection - Read-only links to existing systems
BI Maturity Model Stage 3
A BI System vendor was selected using the following criteria:
- Access can be granted where needed across the Institution
- Security is ensured whenever access is not granted
- The BI system will talk to all of the Institution's data sources (which include both widely used and rarely used commercial systems)
- It is easy for staff in the Institution to set up new links, to drill down to any level of detail, and to change views
- The interactions with the vendor's technical staff were positive
It was interesting that cost was not a problem, since all of the vendors considered came in at similar costs for institution-wide licences.
One vendor met all of the criteria, and provided convincing reference sites and a realistic demonstration (including links to the Institution's more unusual systems). That vendor was selected.
The system selected will be able to receive data from all of the Institution's management systems. (This was demonstrated convincingly during the evaluation phase of BI system selection.) It will not store the data itself (this is not a Data Warehouse model) and it will not write anything to the Institution's management systems.
The update frequency of the dashboards is determined by the update frequency of the Institution's systems: for example, during enrolment, student numbers are updated constantly, and the dashboard shows figures in real time. By contrast, many financial questions (income from catering last month or last term) depend on defined financial reporting periods, and are only updated in those periods.
Initial BI deployment - Senior Administrators and Management
BI Maturity Model Stage 4
Initially the Institution sent about a dozen administrative and IT staff on a 2-day training course. This course, held at the Institution, included making the first links to the Finance, Curriculum and Student systems. These systems were then made available to the senior managers for the Institution.
Following further training courses for the system's administrators (2 days for each for Finance, Curriculum information, ICT and Estates), drill down facilities were added. Staff from the Institution were now skilled and confident at creating new dashboard segments linking different data sources, and allowing drill-down to any level of detail. Senior staff were shown how to customise their own dashboards.
At this point, training and expansion of the BI system paused while the senior staff became used to it and explored its performance.
Expanding BI Use - All Staff, Students, the Public
BI Maturity Model Stage 5
As experience grew, the Institution made the BI system available to more and more staff. Senior staff and administrators in schools and departments were trained to use and modify their own dashboards (by the Institution, who had learned BI and training skills from the vendor). The security model was explained, so they would know which data they could not see and what effect this would have on their dashboard views.
Other staff were given pre-configured dashboards which they could not alter. This was also done for the information offered to students via the Institution's intranet and the information offered to the public via the Institution's website and via kiosks in some public areas.
The Institution has an active environmental programme and monitors water and power use. It generates some solar and wind power and uses grey water. All of these figures are available to the public and to all staff and students.
Forecasting and Planning
BI Maturity Model Stage 6
Finally, with experience of BI across all academic staff and all services the Institution can use their BI system to assist with planning and forecasting. Questions which can be asked relate to changing demand for courses, development of new courses, shared services, and partnerships.
