Ask anyone in this industry for his or hers definition of Business Intelligence and you get a thousand different answers. Is that a bad thing? Not necessarily, but it does imply that we still have not agreed upon some of the basics. It is an interesting paradox that although we are apparently not 100% sure about what Business Intelligence is we do have an opinion on how it should be or where it is going. Over the last couple of years a common understanding of Business Intelligence has been created holding some universal truths firm in its grasp. We decided – just for fun – to turn the BI world as we know it around and see what comes up if we take the opposite view.
Here is our Top 10 of truths:
1. We need one version of the truth.
2. BI projects require a business driven approach.
3. BI development should be done incrementally.
4. BI projects needs high level sponsorship.
5. BI = decision making process.
6. BI is going to be real time (BI 2.0).
7. Information democracy: BI for everyone.
8. BI can not be out sourced.
9. Merging: BI, CI, EI, EPM, CPM, BIM…
10. BI is the answer to everything.
This is my take on things:
10. BI is the answer to everything. Almost every answer creates a new question. Data is often limited, not captured or of low quality. Therefore one can say that BI is not the answer to everything. In the best case scenario it is helpful.
9. Merging: BI, CI, EI, EPM, CPM, BIM. There are so many acronyms and abbreviations that I am not even sure that we can agree on all the two- or three letter words at the beginning of this paragraph. And even if we agree that BI stand for Business Intelligence – I doubt that we can agree on a common definition – let alone talk about merging. To my opinion it is not about merging but a different manifestation where the core or main point may differ per area (focus on strategy, or client, or process…). In each case: it should be information centric.
8. BI cannot be outsourced. A lot of BI professionals tell me that BI cannot be outsourced because of its specific nature (mutual development with end users, incremental approach, and so on). One of our clients has already outsourced its BI environment as long as 8 years ago. The proof of the eating is in the pudding.
7. Information democracy: BI for everyone. Information is critical for almost each person, process or technology. We in the western world produce more data than products. But in order to transform the data into information, knowledge and action you need to add your personal experience, time and effort to the mix. Some people will do this and benefit from it. Some people will not and can live without it. BI is not for everyone but at least for the smart people.
6. BI is going to be real time (BI 2.0). If you view BI as strategic there is no need for real time data. But if you view BI as more operational than it will come in handy. In my view BI supports better decisions by making a strategy, process or objective – accountable, adjustable or adaptive. For accountability you do not need real time, for the other two you do. Therefore my answer would be: it depends.
5.BI = decision making process. This is a tough one. As mentioned previously I think that BI supports better decisions. It is up to the end user if they make use of it. They can make a decision that is in full conflict with the number crunching. Therefore BI is not the same as the decision making process it is just a supporting tool.
4. BI projects need high level sponsorship. The reason for this is that we want to introduce Business Intelligence in the organization. It is all about creating awareness and enthusiasm. So do we really need a high level sponsor or manager for this? Perhaps in a very rigid hierarchic organization you do. But we Dutch do not deal well with a top down approach. The real innovation comes from the work floor, we think. I think organization can benefit much more from a highly committed thought leader.
3. BI development should be done incrementally. What would be the opposite of that? Probably some kind of big bang approach where all the functionality can be delivered in one single release. That sounds nice but often this is not possible for BI. To start with I find that when users start to interact with the new BI system they come up with new insights followed by change requests. A successful incremental approach focuses on the identification and prioritization of the most beneficial increments or slices. This should be based on the priority of the objective (is this a key process, is it in line with our strategy?). Other qualifying criteria can be availability and quality of the data. Each slice should be a complete solution. BI projects – just a real life – should have a first things first approach. Delivering incrementally – keeping the ling term view in mind – allows for faster speed to value. So I agree with this truth.
2. BI projects require a business driven approach. Any BI project or program should have business value. Creating a business intelligence environment without a goal is a mission impossible. We do not need data warehouse projects. We need project that can create more sales or reduce churn. A data warehouse or business intelligence project can support this goal. However I do not think that it is should necessarily be the business user who initiates such a project. IT can play an important role in taking the lead by showing what is possible. This can be done by creating a prototype or have an open environment or playground.
1. We need one version of the truth. Has anyone of you ever played bullshit bingo? It was one of the sources of inspiration for this post. You take 10 of the most used words and cross them out during a meeting when used. Single version would be on the top of that list. I hope that from this day forward no one will ever use this expression again. This was born out of fear for multiple versions in spreadsheets across organization. And frankly, I cannot stand the expression anymore. Also I totally disagree with it. Understanding is always determined by its context! And all communication should differ for each target group. Creating a single version – let alone the office politics involved in that – is an utopia. Instead let us focus on a single version of the facts (data quality, lineage and so on). To me the focus should be on the definitions or Meta data. As long as the definition is clear you can have different version per context.
So are there any ‘truths’ that we missed in our top 10? We would appreciate your suggestions. Also if you disagree with what I said (or agree for that matter) please post a comment. It is a platform for an open discussion about BI – not just a floorshow for out thoughts.