February 27, 2018

We Love Data And Hate Science


To the naive mind collecting data sounds like science. It is not.

A datum is the result of an instance of observation. But observations do not become science until they are made sense of.

People observed the movements of planets for thousands of years. They kept intricate charts. But for all those years no one could explain the seemingly incomprehensible movement of those planets.

And then Copernicus came along and in one simple theory explained what for millennia seemed beyond comprehension. The planets moved the way they did because they were circling the sun, not the earth.

That's what science does. It takes data and makes sense out of it.

Today the advertising industry has unimaginable quantities of data and hardly an ounce of science. Ask any advertising person for data about the Google or Facebook buy they made and you will get reams of papers and stacks of charts and tons of reports. You will get a festival of data.

Then ask that person to name one - just one - major consumer-facing brand of anything that has been built by advertising on Google or Facebook. I promise you, you will get a blank stare. Believe me I've tried it.

And yet, it is almost universally agreed that the primary objective of advertising is to build a successful brand.

So the question is this: If advertising's highest calling is to build a successful brand, and we have no examples of successful brands being built on either Google or Facebook, what is the science behind our obsession with these media?

What is the science behind all the data that has led to the incredible dominance of Google and Facebook if there is not a single instance anyone can find of either of them having achieved the primary goal of advertising?

Where is the science that makes sense of all the data?

Data is just a bunch of bricks laying around. Science takes those bricks and makes a house out of them. Right now we have no house. What data has given the ad industry is mostly just big piles of bricks.

No comments: