Marketing is science

I often say, "the data is neutral". Some new lingo we also love to use is "everything is a bet."

To quote our 90s queen Alanis:

It goes up, we learn.
It goes down, we learn.
It goes down 3 weeks in a row, we learn.
It spikes up, we learn.

None of these things mean you need to change marketing strategy or managers. It means you need to evaluate your process – is it based on a lean framework of incorporating learnings, or is it based on random tests of ego and the distraction of shiny things?

Never believe the first test

Most first tests have "meh" results relative to your hopes and dreams.

It's your first time using the message, meeting the audience, using the channel. Of course it’s not going to go the best it could go! This is where many deem it a failure and stop.

But the best tests are ones where you have repeated it 3 times and improved the variables each time in an effort to find a channel or tactic that outperforms your lowest-performing channel.

Spend 20% testing and focus 80% on building your flywheel and incorporating learnings/optimize as you repeat your best-performing actions. Then, the tests are actually a means to an end – the end being, a performant marketing engine that is predictable, consistent, and reliable. Where 80% of the inputs we can predict the outputs to a degree of confidence, the other 20% can be new bets.

If you're just always testing and assuming if the first test failed “it’s not worth trying again”, then you'll never actually start building a flywheel that can gain momentum because you’ll constantly be starting from scratch.

You’ll get there

Consider that it's almost exactly like a diversified portfolio where the majority of your investments are in vehicles you know perform safely and consistently. But in startups, you have to find what that safe and consistent thing is – and that takes money and patience. Sometimes it requires re-testing things, and with the ad landscape super whack, don’t be afraid to invest in your brand.