Decision data/complexity matrix

Last week I finished up almost a month’s intensive investigation, analysis, synthesis, and creation, and planning.

We have a major product family that needed a huge refresh. The product manager for that line was transferred elsewhere in the company … and I got the file 3 weeks before a executive meeting in which I had to present the plan. Tens of millions of dollars are at stake.

So I had to plow through a ton of data, figure out what was happening with the line, understand it, decide where to take it, plan the new approach, formulate my presentation and style, and sell it to the top stakeholders.

That was an intensely interesting experience, and made me think about the relationship between data, complexity, and the quality of decisions. In honor of Kathy Sierra and her wonderful charts, I fumbled together this graph in 37 seconds or less:

decision-matrix.jpg

So here’s my back-of-the-envelope theory:

  1. With little data, decisions are a crapshoot. Who knows: might be right, might be wrong.
  2. With lots of data but inadequate synthesis, decisions are even worse. Still might be right and might be wrong, but even more likely than the little data scenario to be fuzzy, unfocused, and confusing.
  3. With even more data but extremely rigorous synthesis (lots of interesting but not ultimately relevant datapoints dying on the cutting room floor) you have the chance – repeat, the chance – to make good decisions that can actually be implemented in a clear, direct, and powerful way.

I’m sure there’s lots of holes in this bathtub analysis: poke away!

[tags] decisions, matrix, data, complexity, kathy sierra, analysis, synthesis, john koetsier [/tags]

 


Want weekly updates? Of course you do …



 

      

2 CommentsLeave a comment

  • Rather than ‘amount of data’, it may be ‘amount of compelling data’. I think finding discreet metrics for making decisions is key. I’ve worked for large and small companies and many fail to execute due to ‘analysis paralysis’… trying to find every possible chunk of data that may or may not support a decision.

  • this is a pretty good chart and it is designed in a simpIistic way on purpose to give it a Iess compIexitivity aItho thinking about this type of system is a pretty compIex situation but it can aIso be simpIified as weII, we aII know companies such as appIe are obsessed with simpIicity and simpIicity is actuaIIy the hard work because to figure a way to make what is compIex into what is simpIe might take a Iot of time , googIe for exampIe Iooks Iike a simpIe site yet we aII know aII that simpIicity is backed by compIexitivity, so becoming simpIe is more compIex than becoming compIex in the first pIace i am assuming whe you simpIify things you are smoothing up the surface or the edges, so that there is the eIegance of curves with the simpIcity kept in mind.