Crude measures

“Crude measures of the right things are better than precise measures of the wrong things." — Jim Clemmer, The Clemmer Group

Too often, organizations make massive technological investments in an effort to create the be-all-end-all data platform. Executives imagine a world where any question they ask will be easily answered through this fantastic analytics platform. An analyst will be able to punch up relationships between all sorts of variables, pull out the results in a beautifully formatted chart. Then everyone gets to take off early.

Consider the language used by leading consulting and enterprise software companies:

  • “Only Oracle delivers a complete Business Intelligence (BI) platform, so you can manage, access, and analyze all your business data faster and at a fraction of the cost of alternatives" — Oracle web site
  • “The goal of Integrated Business Intelligence solutions — and the primary benefit — is achieving what we call ‘360-degree insight.’ This means integrating all available types of business information as an aid to decision making… considering all the core elements of your business as a single, integrated whole." — Accenture web site

Wow, that’s good stuff…if you are willing to pay millions of dollars and wait years for these promises of pure business insight. Even so, I would suggest that these massive projects are a risky way to go:

  1. Business directions change. When this killer datawarehouse comes online six or 12 or 18 months after defining the requirements, your strategy is likely to have shifted.
  2. Comprehensive requirements crumble under their own weight. In an effort to answer any possible question, the solution developers consult with all possible stakeholders. After all, it feels right to be inclusive. But if you force people to offer an opinion, they are likely to include low value requirements — whatever jumps to mind. In turn, the difficulty of the enterprise can become exponentially more difficult with each new requirement.
  3. Little effort toward asking the right questions. More data doesn’t itself help the business determine what are the right questions to ask. There is a certain laziness in attempting to get everything, and putting of the effort of evaluating what is really necessary. You could buy an entire Barnes & Noble store and guarantee yourself of finding an interesting book. Or you could do some self-examination, talk to some friends, and just buy the books that you think you’ll like.