Establishing a data culture
AA: How important is it for an organization to have a data culture, and what does a successful one look like?
JTW: I think the data culture question is really hard. It’s easier to engineer software than to engineer data culture. Inherently, some people are just not built this way, and I think I need to acknowledge that. In the distribution of people in the company, there are large populations who are very number- and data-averse, and we need to live with that.
I also think data culture itself has different aspects. Sometimes, we think about it as the ability to interpret the output of a data science project because that seems the most applicable to a data team. But it is a lot of other things, and I think the way we approach it must have different facets.
For instance, someone might never want to see the output of an R model, but they might be OK to understand their own cognitive biases and how their mind plays tricks on them. We have this...