I remember a story a consultant once told me years ago, they embarked on a project that would not end. They were to develop KPI (Key Performance Indicators) for a business.
After 3 months they had a list of over 300, after 6 months it grew to 600. It never shrunk below that number. The reason it comes to mind is that good business intelligence allows for fast insight to drive the behaviour. Makes sense right. When a public company reports information to the street they often have 4-6 major metrics about how their business is doing (same store sales, churn, basket size, ARPU to name a few).
What happens and how do you help a customer when there are a ton of things they want insight into, but the very nature of the approach may lead them into unfathomanable volumes of data or metrics. You know that the business people you are dealing with do care about all of these metrics, that is never a doubt, but that many.
Can you truly drive insight out of a 36k row excel file? Sure if you use that data as a mart to start pivoting against to drive smaller volumes of data. This is when you need trusted relationships and a common vision to align everyone to.
Real insight comes from smaller amounts of data that provides insight. Larger volumes are for supporting, not presenting.
thank your for you explaination , your imformation is very important for me . thank you again.
I remember a similar phenomenon at business school. Professors would ask for key success factors and eager young students would offer up as many as there was time for. In the end of the day there is nothing inherently wrong with talking about a seemingly endless series of success factors as long as you remember that half a dozen or so are KSFs and 594 or so, are Minor Success Factors (MSFs).
600 is too many. But if I have collected many terabytes of data to get results for say 10 KPIs people might have a problem with that. There is an expectation of ROI from the data that is hard to quantify/qualify. Saying we spent $50 million to generate 600 KPIs sounds better than we spent $50 million to generate 10.
How much money do I have to invest per KPI? may be the wrong question to ask. But if it leads to an estimate for an ROI number that will generate funding does it make it a right question?
Once the 600 KPIs are in place the organization may only use 10 of them to gauge corporate performance, improve that performance, or identify events in the data that point out expensive business practice that can be dropped. All this with the end result contributing to the corporate bottom line.
But what if one of the other 590 KPIs provides an insight that makes or saves the company millions because it was created in an EDW environment that was never possible before . Prediction the Black Swan or the turkey event is our KPI Holy Grail.
You are right. I also remember some great reporting applications at Falconbridge, though I did not know as much about that one.
I found the experimentation and value in the work I was involved in for early EIS (executive information systems) at Noranda was very rewarding and insightful.
It is not necessarily about how complex the model is that gives the data, but how much data that comes back. Those early systems had limited availability so you were forced to get to the point and trust the model. So you would fix the model not the volume of data if you were wrong.
Every executive has certain things they watch to determine how well things are going. I recall how what VP of sales measured performance by comparing their selling price of nickel with the price on the London Metal Exchange. That was a great KPI. However not one that was shared with all the executive team.
Alf Powis, a key leader in Noranda Mines, monitored several key numbers based on complex models.
An Executive VP at Inco made his mantra the relentless pursuit of productivity so everyone knew his KPI’s.
These are old examples but indicated how insightful these people were about their business. I think sometimes people need help in simplifying the metrics and working up from details to key metrics.
I think the key is gaining the trust of the key executive so that he can share what he really cares about and what he looks at every day.
We are all challenged to discern the simplicity out of the complexity.