“It’s easy to make bad decisions with data” he said straight off the bat. In today’s tech infected world, it can seem vogue to tout the fact that you’re data-driven. Kind of like it’s a badge of honor that provides an air of superiority over the other ill-informed business heathens. But what is data? What isn’t data? Are the people who call themselves data-driven, actually data-driven underneath? This is something we had to get to the bottom of.
I came across Willem Paling after I saw a presentation he published on how cutting brand search advertising for one of Australia’s largest insurance conglomerates made no difference to the business’ performance. This was something I had also tested personally with client campaigns, but without the same rigor or scale.
So who better to talk about segmentation and targeting than with someone who’s literally built out data infrastructure and carried out experiments himself?
Apart from that, what really sold me was his background. His PHD centered on hard-to-measure variables - check. He was a fan of epistemology - check. He’s worked with both product, services and SaaS companies - check. Large and small businesses, early and late stage. Academia and private enterprise. Multiple and disparate industry sectors - check Strategic planning - execution and measurement - check.
So what you’re about to hear is of course, a discussion about segmentation and targeting. But - to do both well, you really need to have a solid grounding in data analysis. So we talk a lot about how to use data to create value. And on the flipside, how not to use data, in a business context.
What he says is so dense, don’t be alarmed if you have to listen to what he says twice or three times - I know I did - even as I was editing the videos.
Perhaps the biggest takeaway for me from a strategy perspective was how subtle competitive advantages can be these days. He confirms to be that the businesses who are succeeding with are increasingly focused on building more certainty into their decision making via scientific methods - all of which ultimately rely on great data and data analysis.
I’m sure you’ll enjoy doing down this rabbit hole as much as I did.
