March 09, 2016

Data vs Probability


I took my data analyst with me to Las Vegas.

We spent three days at the roulette wheel. We learned a lot of little things. We did a thorough, highly detailed analysis of what went on:
  • We counted how many times red came up and how many times black came up. 
  • We calculated the effect of the croupier spinning the ball left or spinning it right.
  • We calculated how often red came up three times in a row. Then black.
  • We established the effect of the speed of the roll on the number of times it turned up red or black.
  • We computed the number of people over 40 at the table and how that affected the result
  • We counted and analyzed lots of other stuff.
We lost anyway.

Meanwhile the house did no data analysis. But they knew a few big things.
  • They knew there were 18 red, 18 black, and 2 green. 
So they had a 5% advantage on every roll.

Consequently, they never lost.

In marketing and advertising we have two ways of doing things. We can measure every little data point or we can see the big picture.

Small picture marketers know a lot of little things:
  • Who went to what website? 
  • What they searched for.
  • What kind of car they drive.
  • Who their dentist is.
  • Where they bank
  • And tons of other stuff
They create tightly focused advertising and put it in front of a select number of precisely targeted individuals.

On the other hand, big picture marketers know a few big things:
  • People are more likely to buy brands they're familiar with. 
  • People are more likely to buy brands that sound interesting.
  • People are more likely to buy brands they like. 
They work very hard to produce widely appealing materials and put them everywhere.

Then they stand back and let probability do the work.

Why have all the world's leading brands been built by big picture marketers?

Because the more you study data, the more you realize that data is just the residue of probability.

No comments: