The following is a lightly edited transcript of a reply I posted to a question asked on the LinkedIn.com Business Intelligence Group. This was entitled What should companies consider before investing in a BI solution?.
I suggest some of the following:
- What business problems would a BI solution address?
- Within these, what questions do people want to ask and what action will the answers lead to?
- Why can’t these people get the answers today, or – if they can – what is wrong with them (incomplete, inaccurate, not detailed enough etc.)?
- What is the business impact of the lack of these answers (poor decision-making, missed opportunities, inefficient processes, poor monitoring, lack of tools to manage people’s performance)?
- If these questions were to be answered, broadly speaking, which different data sources would need to be brought together (assess different country / divisional systems and different types of systems – sales, Finance, manufacturing, distribution, marketing, complaints, external data, others)?
- How aligned are the various different elements within these (e.g. customer records, products, territories etc.)?
- To what level is the data required to answer the questions identified above captured (are there gaps and does new data need to be entered)?
- How accurate is this data (does it actually reflect business events)?
- What is the overall quantity of both historical and current data that needs to be looked at and how much of this regularly changes?
- How frequently will users need to ask questions and how up-to-date does the answer need to be?