jump to navigation

Data Driven July 23, 2009

Posted by Tim Rodgers in Management & leadership, Process engineering.
Tags: , ,

Probably my favorite mentor and faculty advisor at UC Santa Cruz was Joe Bunnett, now professor emeritus of chemistry and I hope enjoying his semi-retirement in good health. When I was in graduate school there in 1977-1982, Professor Bunnett had a corner office in the Nat Sci II building with beautiful views of the redwood forest that surrounded the campus. Behind his desk was a small, framed, embroidered sampler in Gothic script with the admonition: “Don’t Make Verifiable Assumptions.” I didn’t get it right away, but since then I’ve realized that it made a strong impression on me.

Managers and leaders have to learn to make decisions in ambiguous circumstances with incomplete information. The world keeps turning, and you can’t wait for all the data to come in. The sampler in Professor Bunnett’s office wasn’t saying one shouldn’t make assumptions or base your actions on intuition. It was saying that if an assumption can be verified, then it should be, as soon as it is practical to do so. Failure to verify makes subsequent conclusions and actions inherently flawed and risky. The more basic and fundamental the assumption, the more reason to find a way to verify it.

It’s easy for some to become paralyzed by a lack of data, but it’s also easy to develop a blind spot when it comes to assumptions. A few years ago at HP we convinced ourselves that designing our own software for editing digital photographs and creating DIY projects was worth the development expense because customers who used the applications would use more ink, and of course that meant more revenue. Somehow we never got around to checking that assumption by designing a study to monitor ink usage with and without the software. Part of me thinks we didn’t really want to know the answer  because of the possible implications to the team, but I’m surprised that no one at higher levels in the organization seemed to be interested in verifying that assumption either.

By the way, that’s one reason why newcomers are so valuable to the team: they ask the “dumb” questions that no one else dares to. Challenging assumptions is also one way of inviting opposition (see previous post of July 20).

Verifying assumptions means collecting information; and ideally that should be objective data, not just opinions that lead to more assumptions. Yes, I know you can bias experiments and interpret statistical data to support just about any opinion, but I believe the process of verifying can be made public enough to invite oversight and criticism. What I learned from Professor Bunnett’s sampler has helped me minimize risk in decision making and avoid complacency.

Postscript: At the American Chemical Society’s national meeting in April 2008, Professor Bunnett — who has written technical papers in verse — provided this explanation as to why more papers are not written in that form:

Narration in verse was the rage

In earlier years of our age

Factual statements appeared

Lucid discourse was cheered

Pleonasm was simply an outrage

Presentation written in verse

Is characteristically terse

It takes so much time

To find rhythm and rhyme

To state a truth or deny its inverse.



1. liliancietess - December 11, 2009

I really enjoyed reading your blogpost, keep up making such exciting stuff!

2. Data and Information « Managing in the 2000s - December 6, 2011

[…] about the importance of making decisions based on objective data, see Measuring Performance and Data Driven). I don’t think managers and leaders should wait until all the data is in before making a […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: