Tips and methods of requirement collection for Business Intelligence (BI) and Knowledge Management (KM) projects
1 March 2015
The review is based an article by Linda L Briggs published in TDWI.
In her article, Briggs suggests tips and methodologies for collecting requirements for Business Intelligence based on an interview she conducted with Jonathan Giager, Vice president of Intelligent Solutions. I wish to analyze the recommendations and enlarge their range of appliance. While the tips offered were intended for BI, they are suitable for KM, almost with no change at all.
Hereby are presented the main points stated in the article:
The difference between a KM/BI Project and a traditional project when collecting requirements:
It is more complicated to define a final destination and choose people suitable for involvement in the project.
The requirements appear during the project's run. It is useful to use a prototype for collecting the requirements.
Defining the project is not clear enough, even if the users know exactly what they want; the details are usually unclearly defined in the first stages of implementation.
The project is harder to justify as benefits are in many cases only long term benefits. It's harder to get assigned a budget.
Main properties of KM/BI requirements:
BI requirements share some common properties- they must be as clear as possible, and the main component of the project is not the information supplied, rather the final product, i.e. the business value, which is not supplied by the data/information/knowledge itself, rather by its usage.
Collecting requirements and the importance of the data:
There are two types of organizations-those who have a data manager/knowledge manager and those who don't. Organizations that have a data manager can choose people suitable for involvement in such a project and thus make correct decisions. Organizations that don't have a data manager face an extremely difficult challenge: not making the wrong decisions regarding the data. These organizations should aspire to hire a data manager.
Tips for analysts/KM consultants before collecting requirements:
The most important thing is listening to the business user. If interviews were used for the purpose of requirement collection, it is important to prepare the questions and prepare the interviewers for following questions in case the business user isn't supplying all the required data. The goal is to develop a synergy between the business user who is familiar with the business and the analyst who is familiar with the BI abilities.
The challenge in collecting requirements using interviews
A BI architecture correctly built is flexible to the discovery of further requirements in the future (assuming we reached the correct level of detail). The analyst/consultant's skills are critically important in this case, as he/she can predict requirements the business users cannot.
Dealing with requirements evolving during the project's duration
This is a common phenomenon as analysts/consultants can't predict all requirements as these evolve with usage. A great technique for dealing with a situation like this is using a prototype suitable for the initial requirements, yet equipped with technology that enables inserting further requirements during the project's run.
Knowledge and skills required for requirement collection for a KM/BI project:
Analysts/consultants should know to ask the right questions, differentiate between good and bad information, ask further questions according to the information they receive and know how to interview. Also, they must have a set of analytical & data skills, and obviously a familiarity with BI. The most important knowledge required is business knowledge: what promotes business, business objectives and business environment & competition.
Is finding the right analysts a challenge?
It is a challenge to get the analysts to fully understand the business in order to utilize the content and get the real business value out of it. The problem is that business organizations do not always know how to deal with what the BI/KM can supply, especially in the fields of business analysis and dashboard application.
In conclusion, it's safe to say that BI projects and KM projects are complex right from the start, during the requirements collections phase, and also later on- when running the project and expanding it. The phase of requirements collection itself requires extensive knowledge. Therefore, it is recommended that organizations strongly emphasize choosing analysts/KM consultants that can ask the right questions and realize if the information they received from the business side is sufficient for creating business value and not only pure content. In KM/BI projects, the ultimate value of the project is revealed only after the assimilation of the knowledge (in KM projects) or after the analysis of the data (in BI projects). Therefore, it is sometimes initially difficult to justify the ultimate business value.
Nevertheless organizations should not give up. KM and BI are greatly needed. As difficult as they can initially be- they are later on satisfactory and can produce real business value.