I have a pair of chunky cord trousers in sunset gold that I’m very fond of. They were trendy when NBC aired the first edition of Friends. Until this last winter, it seemed my wife was going to be proven right: they would never be wearable again. But then, suddenly, thick cords were back and I was, for a moment, near the edge of the fashion curve.

It’s a bit like that with Analytics, which was hot ten years ago then somehow disappeared from view. But it’s back with a vengeance. These days, Analytics and BI are de rigeur.

What was the #1 business priority cited by 138 senior outsourcing buyers in a serious market survey published this month? Analytics.

So it’s disturbing that a lot of people continue to look in the wrong place for the answers about Analytics.

They significantly under-estimate the importance of process. They think of Analytics and process as being different worlds; whereas actually they are intimately connected:

  • How could I ever know if a process is working without smart metrics?
  • What’s the point of a metric if it doesn’t measure performance of some aspect of a process, and provide me with intelligence about how to improve it?

Accenture’s marketing push this week Getting Serious About Analytics does recognise that high performance is built upon process ‘discipline’. But Accenture’s analysis of the five most common challenges in Analytics barely mentions process.

I’ve pasted below my comments in sunset gold on the five most common challenges identified by Accenture:

  1.  Focus on the wrong metrics, or too many metrics. Too often, firms fail to map metrics against the key drivers of their business.

Agreed. But how do we know which metrics matter, and which ones are useful leading indicators, not just rear views? Surely only by looking at the end-to-end process…

  1.  Over-reliance on technology. Companies frequently build a data warehouse and then neglect to put analytical tools in the hands of the right people.

Agreed. The stakeholders must define the metrics that matter from the process, after which it’s a selection of the appropriate BI/Analytics platform.

  1.  Inability to select the right data. Managers are presented with too much data of multiple types. Without a proven process for selecting the right data, identifying important patterns is virtually impossible.

Agreed. But you don’t need ‘a process for selecting the right data’. You need to start with a view of the end-to-end operational business process, then define the right metrics by asking questions like: ‘How would we know if we were being successful? What would be the indicator that warns us if we are going off track?’

  1.  One-off, point solutions. Analytical solutions must be connected to business processes or their influence will remain limited.

Agreed. 100%

  1.  Cultural resistance. Many managers—particularly those who came of age 30 to 40 years ago—continue to rely on intuition and judgment. While valuable, these qualities remain limited until combined with the rigor of fact-based analysis.

Ouch! I’m one of those who “came of age 30 to 40 years ago”. Not wholly convinced myself that it’s just us oldies who are leading the cultural resistance to fact-based analysis. But hey, there I go, relying on intuition again 😉