On finding and retaining data talent:
It’s important to have moon shots in your business. You don’t have to make your whole data strategy that ambitious, but you need one or two big, challenging projects to work on that will attract people to the challenge. It might be something they can chip away on or even do a bit of R&D on. Data folks love to not only solve those kinds of puzzles, but also share that knowledge as well. They hang out in those communities and go to meetups, they share what they’re working on with friends and peers, and that means more of them are more likely to end up wanting to work with you, too.
On what data he needs to do his job as CEO:
Lots. I need a healthy mixture of leading and lagging indicators—leading indicators around the top of our sales and marketing funnels and lagging indicators like financials—but, I’m increasingly leaning on data that’s more actionable. That often involves less generic data and more specific. So, for example, if I can break out sales data by geography, industry and transaction types, I can see that deals about capital raise in the U.S. tech industry are converting much faster than the normal timeline. We can take action on that type of data which has a real impact on our business to amplify the good and manage the risk on the others.
On how to spot a good area of investment:
Ideas have to be tested for validation. Even with the smallest test, you can save a lot of time and money.