We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.
Start a Discussion

What's the Key Metric for Measuring BI Performance?

Vote 0 Votes
Taken from an interesting discussion on LinkedIn, what's the key metric to measure the performance of your BI?

7 Replies

| Add a Reply
  • A BI tool or program that does not meet the needs of its end-users will end up as shelf-ware and go the way of Google Wave.

    So, what’s the key method for measuring BI performance in any organization? Simple – The rate of end-user adoption. That is, how many people in an organization who have access to BI software, actually use that software on a regular basis.

    Coupled with this principal, is an equally important measure of performance success – The percentage of an organization that has access to BI solutions.

    If key decision makers in a business environment are unable to access important business data, their ability to make fast and effective judgments on vital operational areas will be significantly impaired.

    End-user adoption is the overriding measure of an organizations’ BI performance.

  • Only one metric matters - business results. Did the BI project improve the quality or timeliness of decisions or not? Did improving those decisions make for better business results?
    Everything else is just fluff.

  • I also think that only possible measure likely be the top line and the bottom line. But unfortunately my reasons aren't very attractive.

    1. Every time I had to do some business analysis, whatever BI or BW were available, I had to use Excel to finally polish my report.
    2. There were just so many BI projects, team A created the dashboard X to manage X for their sake, team B created the dashboard X2 for their sake, and so on. Every team asks field people to fulfill additional details for their sake. It was a hell, wasting time for political battles among management team A, B C, etc.
    3. I knew the field people positively decorated data entries X1 thru X5, and negatively decorated Y8 thru Y12, hence I made phone calls to eliminate such biases, ending up my facts became different from a management dashboard.

    Hence, I cannot tell any specific measures for BI performance, except the top line and the bottom line. Sorry about that.

    Once people stop lying and political battles, then we can discuss practically workable measures, but I'm not so optimistic about it.

    Kengaku, Satoru
    S&S Global Services

  • First of all let’s not forget that BI is NOT a primary requirement for any organization and secondly BI tends to cater to intangible benefits more as it is dependent on other system of records. So given the premise it becomes very difficult to gauge its performance and how effectively it adds value at the end of the day.

    Having said that, to actually measure the performance of BI you need to consider relative metrics – these are derived indicators which are a combination of KPIs using information from various systems of records. These then need to be evaluated which will indicate how much an organization possibly saved due to either timely alerts which helped decision makers to take corrective measures or decisions that were taken by analyzing historical data.

    In reality the performance of BI has more to do with an organization’s profitability by taking into account effective decisions that were taken using a BI tool. And I must also reiterate that a BI tool is just an instrument – the effectiveness or performance depends completely on how it’s being used - there is nothing called BI Out of the Box :-)

  • Is the BI information made available in a timely fashion. Timely is defined by the customer using the information. Outside of that definition I would say the information needs to be made available as quickly as possible to respond to changing market conditions. If I'm a retailer and I can predict a natural disaster in a certain region of the country then I can start shifting my supply-chain towards the disaster before my competitor. If I can get key economic data from the government or other sources and correlate it with my KPI quickly I have a better chance of staying/getting ahead.

    So, the key metric is the time it takes to correlate the data. When do users need it? YESTERDAY!


  • Let me take the analogy of a 'Highway' to illustrate my thoughts on this topic. Typically, a highway is constructed to optimize travel between places in a city. There are 3 broad stages:

    1) Stage 1 - When the highway is being constructed, certain measurements like the safety factor, banking angle at turns, the width of the road, number of lanes etc. are important. Similary during the construction phase of a BI project, the robustness & flexibility of the architecture have to to measured and monitored. Fundamentally, metrics have to answer the question of - 'Is the BI system being built the right way'?

    2) Stage 2 - After the highway is open to public, we want to know whether people really do use the highway? Drilling down further, we want to know how many vehicles ply at what frequency, user profiles, type of vehicles etc. In the BI context, this perspective ties in very well to the 'Rate of end-user adoption' for BI programs as espoused by Glen Rabie above.

    3) Stage 3 - Given that people use the highway, are they better off after the highway was constructed. For example, do people reach their destination faster than before? In the BI context, metrics are required to measure the improvement in the effectiveness of business decisioning after the BI system is put in place.

    Given that there might be multiple BI initiatives in an organization at any point in time with each initiative being in different stages, the calibration technique should allow for multiple metrics to be monitored & acted upon.

  • I dont think there is just the one.. here's a list of 30 Business Performance Indicators You Can (and should) Measure - http://bimehq.com/business-intelligence/30-indicators/

Add a Reply

Recently Commented On

Monthly Archives