KPIs and Data Sciences

By | December 5, 2015

Performance shrinked

Key Performance Indicators (KPIs) are so commonplace in measuring business performance. It is not uncommon to find businesses using variety of KPIs to measure the health of the business. KPIs not only help businesses with something tangible to work with but also sets the direction for the entire organization. Its a known fact that, many executive level bonuses are tied to KPIs. Given that KPIs are so important to businesses, is there a way to improve the all pervasive and critical KPIs using Data sciences? I think there is.

To set the context, lets take an example of a Pizza Delivery and lets assume we are measuring the Order to Delivery time. (BTW, this is a good measure because it’s tangible, easily understood, easily agreed upon and difficult to manipulate. We rarely find such simple yet powerful measures in enterprises.)  One can easily and intuitively come up with variables correlated with the order to delivery time. I am listing some of them:

  1. What day and time was the order placed?
  2. From which location was the order received?
  3. What was ordered? (Size, variety etc.,)
  4. Who was delivering?

While intuition is good and works most of the time, your data may show a different story. At a minimum your data may validate your intuitions. But as a data science professional, I would be looking for more. I could answer questions like:

  1. If I move my location of the pizza store, can my delivery time be improved?
  2. Can I optimize my routes and schedules so that delivery time can be improved?
  3. Can I start a promotion by promising delivery time?
  4. Is my data showing a pattern or behavior which can be used as a competitive advantage?

Delivery Time as a KPI could be linear and can be shown in the visuals below (Very unlikely)

Linear model

3D Delivery time

Or Delivery time could be very complex (very likely) and could be represented something like this (source: www.broadinstitute.org)

Complex

Whether the data is simple or complex, there is technology available today to analyze the data and make business sense of the data . Businesses are always in the pursuit of competitive advantage. Your business data is a gold mine and KPIs are a best place to start with for improving business performance.