BI projects are noble in intent though many managers and technical consultants ignore its implications and prerequisites – data availability and quality, cooperation, maturity, infrastructure, adequate tools and knowledge.
Data availability and quality
The problem with data starts at the source, usually ERP and other IS. In theory they should cover all the requirements existing in an enterprise, though it’s not always possible to rich that state of art. Such needs arrive to be covered by ad-hoc developed tools which often include Excel & MS Access based solutions and whose data are often difficult to integrate.
Data Quality is maybe the most ignored component in the attempt of achieving an effective BI solution. The quality of data is based primarily on the validation implemented in source systems and the mechanisms used to cleanse the data before being reported, and secondarily on the efficiency and effectiveness of existing business processes and best practices.
Data Quality must be guaranteed for accurate decisions, if the quality is not proved, no wonder that users are reluctant of using the reports! Normally the data must be validated, and for this are needed users with adequate knowledge to do that and enough data to back up the BI reports. I would advance here that each BI report needs corresponding supporting OLTP reports, a good synchronization between data sets could offer the base for an adequate validation.
The quality of decisions is based on the degree to which data were understood and presented to decisional factors, though that’s not enough; is need also a complete perspective, and maybe that’s why some professionals prefer to prepare and aggregate data by themselves, this process allowing to get a deeper feeling of what’s happening (even that’s not enough).
A BI initiative doesn’t depends only on the effort of a department (usually IT), but of the business as a whole. Unfortunately the so called partnership is more a theoretical term than a fact, same as managers’ and users involvement.
BI implementations are also dependant on consultants’ skills, the degree to which they understood business’ requirements, team’s cohesion and other project (management) related issues. Everything culminates with the knowledge transfer and training which happens to miss just because of I don’t know what occult reason.
Most of the BI tools available on the market don’t satisfy all business/users’ requirements, even if they excel in some features, they lack in others. I would say that more that one BI tool is needed to cover most of the requirements. When features are not available or they are not handy enough, there is no wonder that users prefer to use tools they already know and see the immediate benefit.
Another important consideration is that BI tools rely on a data model, often inflexible from the point of the information it provides, but also of integrating additional data sets, algorithms and customization. It’s not an accident that customers are finding the concept of data cloud more appealing – it’s a requirement for nowadays business dynamics.
Jumping over the complexity of CMM/CMMI and other such models, I would simplify and say that organizations lack in that knowledge of transforming data in knowledge, understanding it and evolving it further in wisdom and competitive advantage. Most of the fancy words used by Sales People to sell a product don’t become reality over night; of course a BI tool might have the potentiality of fulfilling such goals, though it’s a long way to go until reaching that level.
Infrastructure here refers to human and technical components and the way they interact in getting the job done. Some of the issues were already covered in the previous subsections, while many others can be subject of more elaborated work. I resume in saying that it's not only about “breaking the habits” but aligning people and technologies to the desired level of performance, of retaining and diffusing knowledge. Unfortunately we deal with too much philosophy and less action.