 So I want to unmask what I call qualitative data. And qualitative data, I wanted to pick a kind of cliched term first, data of the heart. I'm calling it data of the heart today because when we talk about human behavior, we've got one of the most complex data sets or data models. That is why people do things, what they do, how they do things. So we can see here from this example, people reading newspaper in the train in the US in the 1950s. But when we think about the why they're doing that, we can see the same kinds of behavior now with people on trains, albeit through mobile devices. And the key thing about this is that oftentimes, technology might not change behavior. In fact, behavior might transcend time and technology. But I still haven't defined what qualitative means for you guys. For us, it means primarily talking to humans. And one of the things that's often leveled against qualitative research is that isn't it less scientific? And why do people ask that? Particularly from the more quantitative areas. It's because, yes, we deal with smaller sample sizes. But I'd argue that qualitative research, for what we do at least, is the only way if we want to understand behavior, if we want to understand what's going on in people's heads, and what's going on in people's hearts, so how they feel about things. But ultimately, our goal is to seek meaning, just like it is in data science, from a big, big complex data set. We want to seek meaning. Our process looks something like this. So again, this is probably quite similar at some level of abstraction to what data scientists do. We're trying to find the research question. We select the most appropriate approach. We gather the data from various data sets. And then we analyze and synthesize that data, or model it. So really, it's about going from research to informed action. In our case, informed action might be a strategy, or a design, or a feature. So let's try and make that a little bit more concrete.