Indicators & KPIs

9 09 2009

In a recent Wired magazine article “American Vice: Mapping the 7 Deadly Sins,” (The original was in the Las Vegas Sun’s One Nation, Seven Sins) a group from KSU students did a great job mapping data geographically.  While in no way is the data perfectly accurate, but in the same way it is a logical indication of behavior.  You can spend time arguing the merit of the work, or spend that same time debating the implications of the information.  Either way, it is a rather entertaining visual display of information.

In the business world, we struggle from trying to be perfect, or perhaps afraid of not being completely accurate.  Indicators do not need to be perfect, they just need to trigger a discussion by highlighting the potential of an issue.  The downside is that when we try, we often try to find indicators for everything (spandex rule) and “a point in every direction is like no point at all” (Harry Nilsson, for those who like eccentic music).  Too many indicators and we spend too much time on data collection and visualiztion with no time for analysis and discussion.  The point is to discuss information.

Business Math – Drivers and Levers

28 04 2009

I once listened to the VP of a services business unit describe the levers available to him to drive the business.  Here is his formula.  While I agree with him about the formula, it was clear that we need to understand the difference between levers and drivers.  


If these were all just levers, would we not be justified in pulling the levers as quickly and as hard as we can?  If we make a small adjustment to the formula, we can take the argument another step farther.


If we breakdown Average Deal Size into its components of hours and rate, we identify the one item that can probably be used as a short term lever – Rate.  We can lower or raise our rates much easier than the other items.  The other items require longer term programs.  We need to train our sales force on new ways to improve winning percentages and shorten sales cycles.  We need to create marketing programs to find new opportunities.

It is one thing to understand the drivers of the business, but we usually need to build programs around the levers to understand how they impact the business and how/when they should be used.  If we are not actively managing the levers, they are really just mathmatical drivers.

Setting Targets

23 04 2009

Setting targets for Performance Indicators should be well thought through. This should not be an exercise in looking at the historical average (unless that is specifically relevant) and then apply 10% as the desired increase. You will want to review history, but you need to understand the goal. It is also important to define the KPI clearly.

For example, let’s use the retail market’s target of sales to sales last year. Retail has traditionally looked at this on a daily basis, as well as rolled up to the week, month, quarter, and year. I have two primary concerns with this:
  • If the weather was bad, we ran a promotion, or some other contributing factor, we may not know it and are really not comparing apples to apples. Additionally, what if last year was really bad? Beating that number doesn’t do much for us. 
  • If we are reviewing this on a daily basis, we loose institutional knowledge due to the repetition. What if we miss a day? Is there any repercussion? What if we miss three days in a row? What if we miss 10 days out of 14? Were there enough days in there of good performance to hide the fact that a trend is occurring?

What would make more sense to me would be to look at this number as a rolling average, or take the total sales for the last 365 days / 365 on a daily basis. Here we can very quickly identify a positive or negative trend, as we don’t have to look at numbers that swing wildly by the day of the week. Instead of talking about  a couple of bad days, we understand that even though we had a couple of bad days, the overall trend is above the goal. We can also integrate our sales goal and show it relative to the trend line.  

Scorecards & Dashboards

16 03 2009

These are two terms that the BI world uses interchangably. The only thing they should have in common is that they both can visually display data.


  • Scorecards are tools that help facilate discussions around strategy and operational performance management. The indicators (KPIs) should foster discussions about corporate direction, resource allocation, priorities, and initiatives. 
  • Dashboards should be used for tactical discussion triggers, like inventory orders, technical support, phone coverage, etc. 

What should be happening with these tools is a far more structured use for each (and throw in reporting as well). All too often these tools are used without discipline which leads to mulitple versions of the truth, lack of focus, red herrings, miscommunication, and ultimately a waste of time and energy.

IT and business users need to work together to better understand what each tool can provide, when that tool will be used, how it will be used, how it will NOT be used, and who should be using them.