The introduction of any new business or information technology process poses risks, but the risks inherent in the implementation of business intelligence (BI) projects are unique and must be handled carefully if they are to be mitigated.
“Business intelligence is a potent business enabler but can lead to resistance from users because of the type of data you are exposing,” says Sean Longhurst, director of EnterpriseWorx, a specialist provider of BI and data management solutions.
“When managers see what information BI can deliver, they are frequently amazed because it provides so much detail. This makes BI its own worst enemy. With the richness of the information available, new users often fear that they may be exposed as poor performers or – in extreme cases – as having committed fraud. And that obviously makes them feel uncomfortable.”
“One of the most important aspects of any project is user acceptance. The users define the success of the project. If they are not on sides it will be an uphill battle. In implementing any new process, there’s likely to be resistance to change, and nowhere is this more true than in the case of BI.”
In the retail industry, for example, a BI system can drill down into specific areas, highlighting where sales are too low or stock is too high, and people are often afraid it will show them up as not doing their job properly.
In addition, people may be used to legacy systems, and feel comfortable with the tools they use, such as Excel spreadsheets. “They often aren’t keen to learn new ways of doing things or may feel that the project is too complicated,” says Longhurst. “That’s another area where BI is killing BI. Users often feel overwhelmed with what we’re presenting to them. BI fails to move forward if users continue to export data to spreadsheets, and remain locked into using Excel software, instead of moving across to the new tools.
“Also, we’re putting information into everyone’s hands. Some people may cling to their own intellectual property and feel that sharing it could jeopardise their jobs.”
As with any IT project, it is important to draw up a project definition document that states the expected outcome and includes an assessment of the risks inherent in implementing the system and ways of mitigating these.
“There is a process flow to installing a BI system,” says Longhurst. “It must be tackled gradually so that users experience small gains quickly and are not overwhelmed by the new processes or by feelings of failure if they battle to get to grips with them immediately.
“Communication and proper training in what BI can achieve is crucial. It’s also important to segment information and deliver it to the right people by implementing the appropriate levels of security. In this way, users are exposed to information that is relevant to their jobs, while senior management have a broader view of the whole organisation, with the ability to drill down into detailed, granular data. This reduces confusion and enables fears to be kept in check.”
Typically 90% of the system is designed to help users, with 10% of information, such as performance BI and job analysis BI, being made available only to senior executives or heads of departments.
Longhurst believes that the executives who commission the BI project and have the final sign-off must be separated from the user community. “That is the only way to ensure that no bias comes into evaluating the outcome – particularly in the early stages – because of possible negative perceptions.