I don’t look my age – but could I also fool the machine?

Piotr Golczyk

I combine marketing with technology, science with intuition and work with fun. piotr @ golczyk.com

I don’t look my age – the vast majority of people are wrong when they try to estimate it. So I decided to check out how good machines would be at it.

How-old.net was the platform I used for testing. Frankly, I did this mainly in the hope that it would stroke my ego a bit. Unfortunately, I was wrong – it was anywhere from 2 to 5 times more precise than people, virtually without fail.

This forced me to think a bit about the new possibilities that technology has created for marketers.

What exactly can a machine tell from your picture?

There is plenty of additional information that can be identified via a camera combined with machine-learning and other algorithms. Examples include age, gender, height, weight, BMI, pulse, mood, ethnic group, and many other traits.

For marketers, it means more data about customers, a better understanding of target groups, and better ad targeting.

What could the future bring? Just imagine combining this data with information about what you’re buying in a store by using a dedicated camera and software. This would tell marketers who exactly buys what, when and how, and big data could easily help find specific patterns in that sea of information. This could be a very useful tool for marketers – without violating personal data protection laws.

The next step will be connecting the dots and finding a way to link offline data with personal online data. In other words – imagine that you’ve just bought a TV in a store, and a few minutes later you see targeted advertisements on your Facebook profile for accessories that perfectly match your new TV.

Sounds like sci-fi? Well, it’s science for sure, and I think it’s also very close to being the future of MarTech and AdTech.