Our Blog

where we write about the things we love

29

Jan

Is Business Intelligence Bad? Looking at BI from the “other” side

As a long-time business intelligence advocate, it has long been my belief that BI is good and BI should be everywhere. But is that truly the case? I look at business intelligence from one side of the coin but each coin has two sides. Can I take a holistic view of business intelligence and continue to assert that BI should be everywhere? Are there instances where – perish the thought – BI is bad?

Business Intelligence

My belief that BI should be everywhere has played a key role in me being where I am today, with Intergen, a Microsoft BI partner; for Microsoft generally does ubiquitous better than anyone else. 

I have worked for Oracle, Brio (who were later acquired by Hyperion) and have done BI consulting for many years. I have worked with people at all levels of organisations, helping to address their need for better, more timely, more accurate information. My belief has been confirmed by customers time and again: BI is good.

Some simple examples of BI goodness include:

  • A large multi-national organisation that was able to learn that many of their expensive marketing activities had no measurable benefit – allowing them to re-allocate millions to other activities and resources
  • A company that was able to evaluate some of their largest consumers of utilities such as electricity and optimise their mix of equipment allowing them to save thousands in overhead costs each month
  • A small company that was able to complete a month-end process in 10% of the time it previously required

In each instance noted above and in many more, BI led to direct, measureable benefit to the company whether it be saving money, better usage of resources, increasing sale, understanding customer behaviour or any of a myriad of other benefits. It is undeniable. It is absolute. Business Intelligence is Good.

Except when it’s not.

How can a BI evangelist, a believer such as myself even entertain thoughts of BI not being good? That too, comes from experience.

There are three categories into which BI is Bad will fit: the process, misalignment and side effects.  It should be noted that one other category could be specified but I do not believe it is an instance of BI is Bad but rather of philosophical difference; that is when people oppose BI due to the fact that it either forces them to change, or they perceive it as a job security threat. Putting that aside, I’ll briefly discuss each of the three categories.

 

The BI Process 

The BI process, properly done, requires a significant time commitment from stakeholders. BI can be viewed as bad when either the time commitment is too high or when the time commitment is too low. When the time commitment is too low the “bad” factor generally is not perceived as a process issue but generally as a misalignment issue. 

The most common reason for excessive time commitment is poor communication. Often people are told there will be a BI project and that they will need to provide input but the time associated with getting their input is not communicated. Most users think it is a simple “tell them what I want and I’m done” meeting and so are not expecting much of a time commitment. In reality, to get a true understanding of needs, in-depth discussions with users is required. Whatever the reason for the excess time usage, for users who still have “regular” jobs to do, they do not view BI as a help but rather as an impediment to their ability to complete their work.

One last comment on the process which again goes to communication, on complex BI solutions there can be a significant time lag between when users state their needs and desires and when a BI solution is deployed and available. If the project timeframe has not been communicated then it can lead to the perception of BI as being something that takes too long to implement and recognise benefit.

 

Misalignment 

This is the most common reason for BI being perceived as bad. Business Intelligence has its name for a reason; it is about the Business. Too often BI becomes WITBI (What IT Believes is Intelligence) or WVBI (What the Vendor Believes is Intelligence). If the business is not sufficiently involved in the specification of needs and requirements (item 1, too little time), they will not get what they want or need.

Because BI plays such an important role in helping analysts and management improve their business, it is recommended that BI be developed in an Agile manner with regular meetings and feedback from users to ensure what they receive is what they need. This is especially important in an instance where BI is new functionality for an organisation or group. Users do not know what they do not know. If they have never had BI in the past, it can be very difficult for them to elucidate their needs and desires in a cohesive manner (at least to developers). Iteratively working with them to refine the needs will add time but will also result in a much better solution, more excitement surrounding BI and higher adoption.

 

Side Effects

BI people often think having access to the data and the ability to analyse is the ultimate BI solution. That view is not shared by all. There is such a thing as too much information. 

You are likely familiar with the phrase “I can’t see the forest for the trees,” indicating you can’t see the big picture because of the myriad of small issues. Giving the wrong person too much information can turn that phrase on its head, “won’t see the trees for the forest”.  Sometimes giving someone too much information detracts from their ability to do their job.

I recall one engagement where we developed an executive dashboard with significant drilling and analysis capabilities. It was used by several executives very effectively. However, one executive allowed himself to get caught up in the ability to look at the data many different ways and failed to perform some other significant aspects of his job. It was clearly a case where by providing the ability to look at the trees, he took his eyes off the forest in order to examine each tree in-depth. His ability to manage the forest was impacted by his newfound ability to manage each tree.

 

So, is BI Bad?

If not properly implemented for the organisation then yes, it can have adverse effects.  Fortunately, there is a solution to prevent (as much as possible) all the items outlined here that can contribute to the impression that BI is bad. 

What is the solution, then, this magic bullet?  It is performing a business focused BI Needs Assessment, often in conjunction with a Data Assessment and then building a BI Roadmap that aligns the BI solution with the business needs and the users within the organisation.

If you are unsure of how to best proceed with BI in your organisation or have had a bad experience with BI in the past, take a step back and think of yourself as a reporter and ask the basic questions: who, what, how, where and why.  Get those properly answered and your BI path has a much greater chance of success. Or drop us a line for a chat. mark.worthern@intergen.co.nz

Posted by: Mark Worthen, Senior Consultant, Enterprise Applications | 29 January 2013

Tags: Business Intelligence, Microsoft, BI, BI implementation advice, BI misalignment, BI side effects, Microsoft BI Partner, The BI process


Blog archive

Stay up to date with all insights from the Intergen blog