February 19, 2020
Online search in healthcare: A conversation with Carrie Liken of Yext
Carrie Liken is the head of industry for healthcare at Yext, a technology company that helps organizations manage their brands online.
For healthcare providers, Yext helps create, manage and analyze information that will show up to the right patients at the right time. According to Yext, most patients search for information about a doctor or provider online, but only a few look at the provider’s website before scheduling an appointment. That means healthcare marketers need to ensure that properties across the web – not just their own site – are in good shape.
Here, Carrie talks to our VP of digital, Reed Smith, about a few keys points healthcare marketers and digital strategists should keep in mind. My favorite note? The one that surprised us the most? The upside down website.
If you like what you’re hearing, be sure to subscribe, rate and review the High Stakes Podcast on Apple podcasts or wherever you’re listening now. You can also read a full transcript of the conversation below the videos.
Over the years, changes in the way people search for information about healthcare providers, as well as changes to search engine algorithms, affect the way hospitals, health systems and clinics should think about their website.
Twenty years ago, search engines would return a list of links. Now, Google wants to keep you on the search page and offers “knowledge cards” so you don’t have to click through. That’s changed the way hospitals, health systems and clinics organize information on their websites and make it accessible to search engines.
Any marketer has to know her audience(s). It’s as true for healthcare as any other industry. And age is just one way to segment an audience (or patient population). The way people use technology at different stages of life has implications for hospitals, health systems and clinics, but the most important thing for healthcare marketers is to make sure they’re thinking about the questions people are asking and the intent behind those questions.
Over the years, people have gone from searching Google for specific words to simply asking search engines full questions. As searches continue to get more complex, as long as they get good results the algorithms will continue to be refined. Liken and Smith point out that patients will be able to get more and more specific information about hospitals, doctors and clinics as this trend continues.
Search engines don’t just aggregate information, they find connections between information through knowledge graphs. For example, Google can make connections between a doctor, the conditions she treats, her location, and who else in the same healthcare system does similar work. Here, Liken and Smith discuss what this means for healthcare marketing content strategies
Reed Smith: Search has changed, obviously, the way people get information over the last 20 years. What does that mean for our website? Because I’ve heard you talk about the website homepage and I know that as a guy that has created a lot of websites, we show people a lot of home page designs. But – and we see it in analytics – that’s not where people are landing anymore. So what does that mean now for our website?
Carrie Liken: Well, there are two things that it means. The first is that we need to think about the website being upside down. So, people come into a website through other mechanisms. They’re not going to the homepage. You’ve seen the data, I’ve seen the data. A lot of people are seeing the data that the URL – the direct load URL – is no longer the case. What people should start to think about is how do you move from having a website to having an intent site.
What do I mean by intent site?
RS: Yeah, unpack that a little bit. What does that mean – an intent site?
CL: If people are looking for information, they’re generally asking questions about something that they want to find. So, if it’s about their healthcare they’re asking questions about their healthcare. What we see is that a question is a signal of an intent. And an intent is, ‘I want to go,’ ‘I want to do,’ ‘I’m going to buy,’ ‘I want to read,’ ‘I want to engage,’ I want to book an appointment.’ Websites now need to be able to deliver answers to those questions without somebody having to come to the website to do that. We call it an intent site. So every single page within the website is a potential mini or micro homepage of the website.
RS: All right. We’ve obviously seen a lot of iterations of Google and SRP or search engine result pages, right? And back in the day, I can remember searching for things and there were no results.
CL: I remember those days.
RS: We’re long past that, certainly. But now I feel like at least we’ve gotten to a place where Google and other search engines certainly don’t want us leaving the result page and so there’s more and more information there – whether it’s star ratings over the years but the knowledge card or whatever that’s called is on the rise. Can you talk a little bit about that? I mean, what does that mean for us as hospitals and especially as content creators where historically if you needed maps and directions you had to click through to our website, and you don’t have to do that now?
CL: Yeah, it’s a good question. The approaches that Google and other search engines are taking now are largely around, like you said, keeping people on Google or keeping people on the search engine results page so they anticipate what it is that you meant by what you are searching to then keep you on that page. So if I’m looking for a person and a person has one of those knowledge cards, Then that person’s phone number shows up on a knowledge card. The star ratings may show up on the knowledge card, a map of where the person is located, the address, places to navigate, pictures, images. So, organizations need to start thinking about getting more of that information to those places, and how do you ensure that that information is correct and accurate?
The other piece to think about too is that there was a recent study that came out about these zero click searches. People will search on Google and they’ll search on Bing and they’ll search on a lot of different search engines and find exactly what they need without ever having to click. So, technically speaking these search engines are becoming largely more of the website like the home page itself. And organizations need to start thinking about understanding that data and how to get that data out. But then also like the mini homepage concept of making sure that the website is not just a single piece of information that’s out there about an organization, but all of the different pages within the website. How do you make sure that that shows up in the search engine results page in case somebody is not looking at a knowledge card but they want to click through and find the right information and they’re coming in through website via other means, not through the direct homepage itself.
RS: All right. So, we see the need for organizations to create this idea of a knowledge graph or associating content to ultimately help people find answers that they need. That is what Google has done, search engines have done is you’ve mentioned. What does that mean though for our content creation strategies and things like that? Does that take us in a different direction or what does that mean…?
CL: Right. There are two ways of looking at the knowledge graph: You can look at the information that lives in the knowledge graph, and that is basically information that’s connected to other pieces of information. So, a doctor works in a particular location, specializes in certain conditions, accepts certain insurance. But what if that doctor also published certain pieces of research? What if that doctor also treats certain conditions where then you can build content off of those conditions? You could think about the doctor’s information in the graph and the relationships among all of those things if somebody’s searching for, ‘does this doctor treat this particular condition and accept this insurance?’ But if somebody wants to know more about a particular condition, how do you attach that condition to that doctor? Or, if somebody is researching the condition then who treats those conditions – maybe it’s five different doctors within the system. The other way of thinking about it is taking a look at the search data that an organization already has and starting to see whether people are starting to search for things more now than they had been last month or two years ago or 12 months ago. And if you start to see trends there and you realize that you don’t have the content, then that’s an opportunity to create the content because maybe there’s something there that people are being triggered by externally and they want to find out more. You can add that to the graph and then that should generate more information and more results when people are asking the questions.
RS: It’s what we historically did with focus groups, right? Like, in-person focus groups. We asked a bunch of questions, people gave us a bunch of answers. Well what’s happening now is we just don’t think of it that way. People are telling us stuff just continually, and we should use that to inform the content that we’re creating or how we’re structuring or building websites, campaigns we’re running, things like that. So that makes a lot of sense.
CL: Yes, just treat search as your focus group and you’ll get automatic answers all the time.
RS: Obviously in healthcare we have a number of different audiences that could be physicians, it could be employees, but in most cases, we’re talking about patients, potential patients, things like that from an acquisition strategy.
Should we think about search as it relates generationally any different? You know, I’ve heard you talk about – and it was really funny because we used to talk about this with social media and everybody thought that that was for the young kids, you know, and you would look at Facebook and realize that the 55-plus was the fastest growing group for most of these hospitals.
So, what are we seeing in search? Kind of along those lines as it relates to generations? Anything in particular?
CL: We’re not necessarily seeing age group-related search differences. Everyone is going to search. I’m just going to put it out there. I have no data to actually back this up, but everyone’s going to search to ask a question or to find information.
Whether you’re 82 and you’re using an iPhone or you’re 65 and you’re using an iPhone or an iPad or you’re 21 and you’re also using a mobile device and maybe a voice device because it exists somewhere in your space. So, no matter what, people are going in there asking these questions. You can’t necessarily say that the 82-year-old is asking anything differently than the 21-year-old. That 21-year-old might be consulting a piece of technology far more than somebody who is older.
What we do see though, is some technologies are being adopted a little bit differently. We saw in our data over the last 12 months or so is that there was a 29% increase every year in people age 65 and older on using voice devices to find healthcare information. Okay, so there are ways that we can understand how people are using the technology but search in general is all about, just think about the questions people are asking. Think about the topics that they could be wanting to find more information on. Think about the intent behind it and then being able to deliver the information rather than figure out the age brackets and what is specific to those brackets.
RS: You mentioned voice which I think is interesting because now we’re going from, we wanted to be on the first page of Google to the top half to the first three to like now it’s like the one. It’s like just one response. And that completely I think changes what we’ve done historically especially as it relates to metadata and some of that stuff that we’re trying to get to the top of that first page.
CL: Right. There’s a whole technology behind how to try to get to that position Zero. It’s about structuring questions on the landing page itself in a certain way, about total number of sentences that answer the questions. It’s about schema tagging on the back end of the page. It’s about trying to figure out, how do I bullet the results that I have in the answers to the questions so that Google can pick it up. So, there are a lot of different ways of now thinking about how do to structure the website and the individual web pages within the site itself to be able to achieve that position zero. And you can’t guarantee it. Google could change it tomorrow. It’s all, we’re all dependent on Google and the algorithm.
RS: So, you know, we’re talking about this, and I’ve heard you talk about the move from keywords to questions, and people are asking questions, obviously online. And those Google queries are getting longer and longer, especially on average.
Right? How should we start thinking about that as it relates to consumer-based service lines in a lot of cases? Are people searching in just different ways? Are they asking differently when we’ve moved from one or two words to the ‘near me’ searches to now these long queries? Where does that kind of progress, where does that go?
CL: Where does it go… Well, it goes from historically ‘doctor near me’ to, ‘cardiologist near me’ to, ‘cardiologist who treats mitral valve repair’ to, ‘who is the best cardiologist near me who treats mitral valve repair’ to, ‘who is the best cardiologist near me who accepts blue cross who is accepting patients on Tuesday at 2:00 PM’ to… who knows where it could go. But what we see in the search data is every single one of those questions. So, I had mentioned in the past that we have seen search queries that have been as long as 136 words long. Very very very very very targeted searches. And people know that they can search today and find the answer that they need, and they know that if they get more and more granular, they should be able to get a better search result. What we’ll see is more searches will become more complex. As long as people are getting good results it’ll keep feeding the system.
RS: Just keep refining.