This is about analytics.

This is about day to day life.

This is also about how the things should not being done.

Incomp… lete?

I have had a friend at home yesterday and the conversation besides other things went on the matter of work. She works in the health area with disabled kids and of course she has to deal with Excel, filling out the metrics on the kids progress.

Now to the “funny stuff”: she has 31 kids to work with, giving each of the kids 30 minutes per Month of her attention. Yeah, I can imagine the readers emotions – thinking that it takes 30 minutes to adjust to another human being… Meaning that almost no meaningful work can be done.

Now, she has a colleague who works with just 4 kids per month, with the same hour investment in total. Hahaha … Imagine, this colleague evaluations looks much better! Damn, of course they do – because the amount of time spent per person looks quite different and the number of project switches is much smaller, so the colleague can focus on the few essential projects.

Doing analytics on the wrong and incomplete data is bad. It can lead to very bad results indeed, because the improvements per kid in those 2 cases are simply not comparable. Period.

Incomp… etent?

If you are using just the basic math like (2+2+2)/3 – well, breaking news! You might be doing something wrong! Life is not 2+2, otherwise 9 pregnant women would deliver a baby in 1 month. Otherwise a team of 10 professionals would produce 10 times more on a project.

I laugh people in the face whenever I hear Human Resources about the normalisation and the infamous 20-60-20 principle, meaning that of any given enterprise 20% are the top-performers, 60% are doing OK and 20% potentially should be fired.

For the Basic Incompetence sake, does this mean that if you build a top team out of the best 5 professionals, then after a year of amazing work you should fire 1 of them and promote 1 of them to be the boss ? What criteria will you use ? What if you hurt all the team by firing that 1 member who was simply the soul of the team ? How can you measure their friendship ? How can you measure the friendship of their partners, kids & associates ?

You gotta be pretty incompetent to base real life decision on simple mathematics – its like running in the total darkness in the woods.

Incomp… rehensible?

Analytics?

You can’t measure people, you can’t measure emotions – no matter how hard you try.

Prediction will fail. Again & again & again.

Because the truth is – if everything is being measured and predicted, then life won’t be interesting enough to live.

We, humans, need surprises, we are unpredictable. Or at least we used to be.

Predict that I can mix a song in an album of Britney Spears and Metallica, with Pavarotti and Jean-Michelle Jarre, plus Duke Ellington and Loreena McKennit.

Predict that some people will totally change their mind based on some external factors.

Predict that life can be created not only on the base of the plan.

Predict not with 36.5454656%, but with a certainty.

Predict that people have feelings and if you treat them badly based of some average statistics – they will drop your brand, change the contractor, move to a different country.

Measure what you can, but be very sceptical about what that means.

Very, very sceptical.