Picking up from Tuesday’s post about the Three Big Lies in Data

I give you…  the World’s Most Honest Scorecard!  


In Moments of Honest Reflection, Most Organizations Would Acknowledge that
this Describes their Relationship with Data

The Most Important Numbers Generally Don’t Exist


The two graphics above kinda “steal the thunder” of this point, but I’ll elaborate anyway.  One of the grandest surprises for me, after leaving Microsoft, is just how blind the world is to its own data.


I think the biggest problem in the world of data today is STILL that we simply DO NOT HAVE the “numbers” we need.  Call them metrics, values, KPI’s, whatever – they are numbers, about your business, that would inform/drive improvement.

Pause and let that sink in for a moment, because a) I am 100% confident in saying it   and b) I think it’s time we admit that to ourselves.  How can we expect tools vendors (or the Gartners of the world) to be talking about that if WE ourselves aren’t willing to admit it?

It’s OK, Everyone.  Let’s Have a Group Hug.

imageNo one should feel guilty about this hard truth.  I’ve interacted with hundreds of companies, in one way or another, since leaving MS.  And NONE of them had all of their important numbers.  (Yes, absolutely – some were farther along than others, but to even having 40% of your “needed” metrics would earn a gold medal at the Olympics).

In fact, that is the reason clients come to us in the first place – they have data, and they need help turning it into information.  So you’re not alone in this, far from it.

We are just NOW reaching a crucial intersection of tools and awareness to change this reality.  There was, quite seriously, no way to avoid this “we lack our numbers” reality until now.

Where does this lead?  It all leads back to an “unsexy” term.

We need models.  And no, not Giselle Bundchen.  I mean “engines that turn data into information, and that can quickly/flexibly answer varied questions about that data.”

Furthermore, we need models that can be built QUICKLY, and quickly adapted to changing needs and awareness.  SSAS Multi-Dimensional Cube technology – the forerunner of Power Pivot – was/is great at modeling.  But in practice, it is too cumbersome to pass the “quick/flexible” test.

Next Week

I will come back to this with some tangible advice on how to actually evaluate tool sets.  Things you should ask vendors, things you should ask yourself, and principles you should consider.