Update 1/22/13: A lot has changed with Facebook and Google in the last couple of years. Please note that much of this post is now out of date.
In part 1 of this patient recruitment series, I went over introductory targeting techniques for research professionals new to Facebook advertising. I suggest you read part 1 even if you are experienced with Facebook ads since it forms the basis of what I’ll talk about here. Check it out:
In the first portion of this series, we relied on very basic tactics to guide our Facebook targeting. With the additional data that part 2 provides, we’ll improve our patient recruitment targeting significantly. If you read through the end of this post, this data will give us some interesting revelations about the interests of diabetes patients.
It’s certainly no accident that two of the most successful companies, Google and Facebook, are data hoarders. In the information age we live in, data is extremely valuable. For better or worse, both of these companies have amassed a ridiculous amount of data simply by leveraging the human love of FREE to their advantage.
But Google and Facebook are very different in the manner they collect their data. In the most simple terms, their approaches can be contrasted as search vs. social. Each of these approaches has their respective strengths. In this post we are going to leverage Google’s search strengths to breathe new data into our patient recruitment targeting. But first, we need to locate our patients.
Locate Your Patients
Lets start with a Google search and see what diabetes websites appear on top of the natural search results. Here are the first few listings:
As you assess these websites, it is important to distinguish between sites containing general disease information and sites appealing to visitors who have (or are close to someone with) the disease. We are not interested in people with a general curiosity about medical information. Find popular websites for people seeking information about treatment, symptoms, management and other topics of interest to those directly affected by diabetes. These websites are frequented by our target audience, which is why they will be useful in the next section.
For example, Wikipedia is not an ideal option because visitors to this page are often looking for general information. The WebMD page with a diabetes subdomain, on the other hand, will have a strong concentration of visitors in our target audience. Not only are visitors of this website likely to be affected by diabetes, but they will probably be receptive to our patient recruitment ad since they are already seeking health information online.
Learning About Your Patient
Now that we have selected a website frequently visited by our target audience, it’s time to hop over to Google Ad Planner. Make sure you have the “research” and the “search by site” tab selected, paste the domain in the box, and hit enter. This tutorial is US-centric, but Ad Planner has data for other countries as well.
Google gives us all sorts of interesting information about people who visit the WebMD diabetes website. If we’ve done a good job selecting a website frequented by our target audience, this information is exactly what we can use for patient recruitment targeting. Google Ad Planner provides us with demographic data, including age, gender, education, and household income.
A quick check of the CDC website tells me that diabetes affects men and women (relatively) equally, but the majority of visitors to the WebMD diabetes page are women. If you are knowledgeable about demographics for the typical “epatient,” this discrepancy should not be surprising. Women are more likely to look for health information online, as are those with at least some college. Another explanation for this discrepancy is that women with diabetes tend to have a lower quality of life than their male counterparts, and thus have more motivation to seek information. Regardless of the reason, this discrepancy is important to note.
Facebook offers targeting options for both gender and educational level, so this demographic information is certainly something we could use for our patient recruitment campaign. But just because we can doesn’t mean we should. Always consider how targeting might influence the data integrity of your study. In this case, limiting your targeting to women and excluding men would yield a better CTR. However, you don’t want all of your patients to be women, since diabetes affects both genders fairly equally. But knowledge of this gender dynamic can still be of use.
Alternatively, the information yielded by Google Ad Planner can inform our campaign optimization practices. Since a gender disparity exists between patients with diabetes and those seeking diabetes information online, it would be wise to separate our campaigns by gender. In all likelihood, these two groups will respond to ads and landing pages very differently, so they should be treated as separate Facebook campaigns.
Google also provides information about our target audience’s interests.
As you can see, Google provides an affinity score. According to Google, the affinity score “estimates how many times more likely you are to reach an audience who visits a specific site or searches for specific keywords versus an audience on the internet overall.”
Our diabetes audience has an affinity score over 3 for vegetarian cuisine, soups and stews, fuel economy and gas prices, and small kitchen appliances. The food and small appliance interests aren’t shocking given the importance of diet to diabetes management.
But the interest in fuel economy and gas prices is not the first interest that would have occurred to me. Apparently United States diabetics are interested in this topic approximately 3x more than the general population. I haven’t checked, but I would bet there are no shortage of Facebook Groups complaining about gas prices, which provides us with more people to target.
All interests listed by Google can provide ideas for interest targeting on Facebook. Pay particular attention to those interests with the highest affinity scores. With new interest areas to use, we would want to continue research with these topics in mind and brainstorm related keywords to use for patient recruitment targeting.
As an aside, please note an important consideration when you use the “search by site” tab. Ad Planner only pulls data specific to the domain name and not individual pages on the domain. For example, we used the domain diabetes.webmd.com. In this case, WebMD used a subdomain for its diabetes page so we were able to use this domain in Ad Planner to get diabetes-related results. Had the webpage been webmd.com/diabetes, Ad Planner would have pulled data for the entire WebMD domain rather than the diabetes page on that domain. Obviously, we do not want data for the entire WebMD site.
I Got Skills. They’re Multiplying.
Please forgive me for the Grease reference. I’m not sure I can forgive myself for it. Hopefully the information I present in this last section will win your forgiveness. Hopefully.
The techniques presented in part 1 and 2 of this series are just the beginning of Facebook targeting strategies. I could explore many more targeting avenues for our hypothetical diabetes study, but I don’t want this post to get too long or confusing. And the best way to master targeting is to practice. If you really want to learn more, your time is best spent experimenting with Facebook’s targeting and tracking your results.
My other patient recruitment blog posts often emphasize testing and tracking, and this post is no exception. Testing and tracking are crucial elements to online advertising success. Once you have the basic tools for Facebook targeting, keep testing and tracking.
With that said, I want to highlight additional resources that will be of use. If you have tried Google Ad Planner, you may have noticed that I didn’t discuss all of the information offered by the “search by site” tab. Information like “sites also visited” and “keywords searched for” will also aid in the development of Facebook targeting strategies.
In addition to allowing searches by website, Google Ad Planner provides searches by audience. I’m not going to go into all of the features in this tab, but they are plentiful and powerful. Experimenting with the “search by audience” tab would be time well spent.
Several websites offer services similar (or complementary) to Google Ad Planner. The best to start with are Compete, Quantcast, and Alexa. Each of these websites has unique strengths/weaknesses and data collection methods. None are 100% accurate, including Ad Planner. For this reason, you should use several resources and research techniques to plan your campaign. This lack of accuracy also highlights why testing and tracking are so important.
As I mentioned in the previous post, we did not have to consider real life factors in our hypothetical study. In reality, you will have study-related considerations like inclusion/exclusion criteria that will effect your targeting. I can’t possibly go over every caveat or complication that can arise as you plan your campaign. However, this series is a good foundation to build on for future Facebook campaigns. Good luck on your patient recruitment campaign!