Brian Cusack: Google’s New Study on Patient Health Research Behavior

Health 2.0 Co-Chairman Indu Subaiya spoke with Google’s Head of Industry (Healthcare Services) Brian Cusack about his company’s new research around clickstream data analysis. He also explained the proven validity of the clickstream analysis results, as shown by the similarity of the results to concurrent survey data.

Google’s new service may be the solution for many health tech quandaries, including the question of how to direct patients’ online search into clinical care. Brian will share more at the Sixth Annual Fall Conference on the Big Data Meets Real [time] People panel, Monday, October 8th. More interviews with Fall Conference speakers can be found here.

Indu Subaiya: So to start off, can you tell us a little bit about the study design?

Brian Cusack: Sure. There are two components in the study. There was the click path or clickstream portion. So, we partnered with Compete — do you know Compete at all?

Indu Subaiya: No, I don’t.

Brian Cusack: Okay, so they are a digital research firm — kind of like a Nielsen for the web world. That’s oversimplifying it, but they have a panel of two million US citizens essentially. And they essentially track their activity on the web.

So we partnered with them and looked at a percentage of those two million that are identified as doing health research on the web. And we were able to follow them through to conversation. I think we looked at six different brands in particular, Mayo Clinic, Cleveland Clinic, Cancer Treatment Centers of America, and Anderson. And so what we potentially did was we kind of reverse engineered that clickstream behavior from conversation. That is, conversation could be anything like, “they made an appointment, to they contacted a doctor, or contacted a hospital,” right? So it doesn’t absolutely mean they became a patient but they took that action to becoming a patient of a hospital.

And then essentially, we backed out from there, “what was the path they took from right there?” That’s the big data clickstream component of this.

Then we looked at that audience and at a representative group that reflected geography and age and all types of demographics. We did a survey against those people and surveyed them before other behavioral things.

So, not just if you chose a hospital, but what are the primary motivations for choosing the hospital? Is it reputation? Is it that they’re technologically advanced? If you used video, what types of videos were most compelling for you?

We did a classic study, a survey against a subset of the clickstream data audience that I described to you.

So the data is highly analytical clickstream data, and then also the still statistically significant, but a little softer survey data, is what we have for information.

Indu Subaiya: And just to clarify, when you say they called or made an appointment with the facility, are you saying that you attracted them because they did that behavior online or they went off-line to make the connection with the provider?

Brian Cusack: Could have been either.

Indu Subaiya: And regarding the multiple devices and platforms, how did the analysis extend pass what they are doing? Purely through clicks? How do you do that analysis?

Brian Cusack: Obviously, the activity on smartphones is trackable in this way. Activity on tablets is trackable this way. Digital video is trackable in this way. And then some of it is also captured by the survey.

Indu Subaiya: Yeah, and that’s fascinating. So what is, then, the response of the partners with insights like this and where?

Brian Cusack: We haven’t gone to the market with this yet. So when we show this to your audience in a few weeks, it’s going to be really fresh.

And in fact one of the things I just learned today, that I am hoping that – that actually might be very powerful is that we are going to try to create a really interesting infographic around a lot of this information. So I could have something that doesn’t even exist right now, in an infographic in a few weeks.

So your question is, how are the partners reacting? Don’t know yet, because we haven’t showed it to anybody.

Indu Subaiya: Wow! This is great.

Brian Cusack: This is absolutely fresh. I was just sitting in a meeting and we were digesting, and cutting, and understanding where the data fits within the area. And we’ll be doing it again. So this is totally fresh.

What we’ve been doing in the field really to this point is telling the same kind of story but without the statistical data to back it up. Now we have the backup, and now know in fact there is a path, there is a journey that a patient follows…

Today-more than ever-that path involves the digital experience because it empowers people more, it gives them access to more information, the whole concept of video layers … a whole other piece of it.

Now, patients can become active partners in their own treatment.

We’ve been telling that story without data for about the last six months, and some people really grasp it and say “we get it and we like it; let’s figure out how to leverage it.” But now we’re going to have the data that specifically supports the story of a patient, and be able to predict the result.

As a result, people will be creating digital video strategy for their hospital sites, rather than just advertising strategy. They’ll be able to compare how they integrate what they do on their site from an information perspective, but also from a video perspective, and yes, of course search as well-the beginning point of the whole process.

Indu Subaiya: I think it’s very, very interesting. And the more you’re able to actually get other data points from the consumers themselves, the more targeted these strategies can become, because maybe somebody searching for information around a particular condition behaves differently than another.

Brian Cusack: You are absolutely right. If we were talking about the automotive market, then we’d be searching for four door sedans versus sports cars, and you can act on that. However, due to privacy concerns, we’ve got to be really careful about how we act on searches against certain symptoms; very, very careful.

Indu Subaiya: That sort of tension around what the constraints are around this type of analysis, I think that would be an interesting theme to explore, but this is actually fascinating, Brian.

Fantastic, well thank you so much for taking the time to talk to us. We look forward to seeing you in a few weeks.

Brian Cusack: I’m looking forward to it as well.

  • Sherry Reynolds

    Does this imply that they are tracking if people are actually calling a doctor’s office?