When Kolker became CDO at Seattle Children's, "I can tell you that for the first years it was very tough; we were not sure what to do, how to position ourselves."

But over the years, some best practices have emerged.

"You need to be aligned with both business and IT," says Kolker. "Business first, IT second."

In obvious ways, technology can make the job of a data scientist much easier, he adds, but without the right strategy in place it can also make it much more complicated.

"In our place we're working with the IT team very closely," says Kolker. "And it helps a lot because you have to navigate all these different systems.

"Compared to the big boys – I'm looking out my window at Amazon – or Google or Microsoft, we're not really big data. But we're complex data. And without IT that's very challenging."

But the past few years have seen "a lot of positive changes happening," he says, especially when it comes to Seattle Children's "current focus on utilizing benchmarking, to improve quality and safety."

Any advice for other organizations looking to dip their toes in the data-driven waters?

"They need to decide that they're going to be open-minded and are going to think about what are the important priorities for them right now," says Kolker. "When they figure out that one, they immediately can see potential wins and go along those lines in order to get to the key priorities.

"Then you build this multidisciplinary team of practitioners, of leaders, of managers, and also hopefully people who are inside and deal with data," he adds. If not, you can bring somebody from outside who has experience with these challenges and opportunities."

That "people angle" shouldn't be overlooked, he says.

"People talk about technology, and we need to have it. Data science, analytics? Absolutely. Business practices? Definitely. But still the major focus is people – who are going to make decisions or not, make interventions or not. The whole focus of what we do is to help people make better, data-driven decisions."

Once those strategies start to bear fruit – even modestly – with clinical or financial gains, targeted use of data can become addictive, says Kolker: "When you see the first successes, success breeds success."