The Measurement Paradox

John Miller, SPCT

What is the Measurement Paradox? 

We measure what is not valuable.

We do not measure that which is valuable.

Seems kind of odd – doesn’t it.

Douglas Hubbard states this (and I paraphrase):

We measure what is simple to measure, but valueless in that it does not provide us with information having a high decision analysis value.  We choose to not measure what is valuable (it has a high decision analysis value), but is difficult to measure.  (How to Measure Anything: Finding the Value of Intangibles in Business, Douglas Hubbard)

In addition, when we have no information, we believe we need much information to make a better-informed decision.

Turns out the opposite is true. When we know little, we only need a little information to be better informed. When we know much (or we think we do) we need much more information to make a more informed decision. Read what he says about Emily and Fermi.

These are all part of the Measurement Paradox.

In addition, to me, his discussions on measurement also help me describe how this is a key factor on why 1) teams can get really good at estimating stories in 3-5 sprints (iterations); and 2) why Product Managers (along with the Program Team) can get really good at estimating Features (SAFe® usage of the word) in 3-5 PIs. Noticed I said can. They have to choose to do the estimation technique! They have to care. Or, as Yoda states, "or not".

Hubbard also states, when the estimator chooses not to care, their estimation never gets better!

Does this sound like any teams you know regarding story estimates?

When there is a team whose GACM is permanently set to zero or near zero they will claim they can not estimate.

In Hubbard's workshops he helps people learn how to estimate in the 100's of millions of dollars.

And the teams you are coaching are pushing back with

"We can't estimate this" 

"We don't know"

OR WHATEVER!

(btw - do the above with a very whiny voice and you'll hear your teams and their reasons).

In his book “How to Measure Anything” Hubbard describes how we get much better at estimating (his word is calibrated) after just 5 exercises (uses) of the same estimation system.

Kind of interesting isn’t it.

Yet, in my experience, talking about estimation of stories always gets someone riled up. Kind of like a religious war.

The Leadership you are coaching will want to understand this. Talk to them, help them understand.

Some managers just want hundreds of measures so they feel good having meetings and managing. Let them self-select out.

Want to learn more? Read Hubbard’s book “How to Measure Anything”.

Need to talk about it? Let me know, drop me a note.  John@LeaningAgile.com

Mentoring for a beer.

Kind of like me listening like a bartender does while you commiserate about things!

jm