Editor's note: Lorin Drake is vice president, consulting services, at Schwartz Research and Consulting Inc., Tampa, Fla. He can be reached at 813-207-0332 or at lorin@schwartzconsulting.com. This article appeared in the February 25, 2013, edition of Quirk's e-newsletter.
If you follow social media, you're probably aware of Facebook's most recent announcement. To summarize, the data wizards at Facebook have devised a way to catalog, find and organize the legions of information Facebook keeps on its members. Facebook is calling this new feature Graph Search, a derivative of what Mark Zuckerberg calls the world of interconnected people online: the Social Graph.
At face value, this is actually a fascinating and ambitious undertaking and the implications and advantages are still being uncovered. After all, there is no company on Earth that has more personal data on us than Facebook. Every day, its over-one-billion users pour their hearts out on the social network, discussing their triumphs and small personal victories and their setbacks and disappointments. All of this information-sharing is done voluntarily, happily and without any moderators, questionnaires, surveys or forms to fill out. It's the world's most passive - and most powerful - data collection engine.
A world of new possibilities
Marketers have recognized the limitless potential of Facebook and have also embraced the social network by attempting to create relevant branded pages and content-marketing strategies, as well as highly-targeted advertising placement. But Graph Search introduces a world of new possibilities. It is also notable since it's a very large leap forward in search capabilities without any involvement or help from Google. Facebook uses a sophisticated and proprietary system to search its own user database. And where it needs to search the Web, it uses Microsoft's search engine Bing. This type of powerful search capability without intervention from Google is actually a significant development. It puts Google in Facebook's crosshairs and cuts Google off from what has become a significant part of the Internet: Facebook's exclusive and coveted blue-walled garden.
On January 31, Ad Age wrote the following about Graph Search: "As more and more consumers use Graph Search, Facebook will become the No. 1 rival to Google, collecting a tremendous amount of intent-based data to improve ad performance."
A quick test-drive
Before I discuss what I believe are some very exciting implications and possibilities for the marketing research industry, let's take Graph Search for a quick test-drive. Not everyone has Graph Search yet, as Facebook is rolling it out gradually, but if you do, you can follow along.
Let's say I want to find out which of my Facebook friends like sushi. I simply go to my Facebook page and type in "My friends who like sushi" and, voila, I get a list of friends who like sushi. I've blurred out their pictures and last names to protect their identity but I think you get the idea (Figure 1). This is a fairly simple search and it generates a fairly simple output. Four of my Facebook friends also like sushi. That's enough of a group to invite out for a nice sushi dinner.
Figure 1
But if you look at the column on the right, you'll notice that this search can be further refined. Graph Search allows you to fine-tune your search by gender, relationship, employer, current city, hometown, school, friendship and other Likes. This is where Facebook starts to look like a database and data people like me start to get excited.
This is also where we can begin to see the application for marketing research.
Let's say we are doing a focus group study with people who own dogs, who feed their dogs Brand X and live in Tampa, Fla. I'm actually using a real-life case study from our firm but am not identifying the client brand.
A quick Graph Search reveals that there are more than 100 people who meet these criteria (Figure 2), certainly plenty of people to recruit a few focus groups and possibly also enough to conduct a small quant study.
Figure 2
Where it gets interesting is how granular Graph Search allows you to get. Not only can I find dog owners who feed their dogs Brand X in Tampa, I can also extend this search to see if they shop at Publix, PetSmart or Target.
The disruptive part
Are you starting to see the possibilities? One thing to point out is that this method of finding and qualifying potential respondents turns the traditional and current marketing research model on its head. And that's the disruptive part.
Ever since marketing research migrated from door-to-door to phone/mail to online, the process of collecting data from respondents has operated under the same buyer-seller business model:
- Buyers are those who seek data, insights and information from respondents and are willing and able to pay for this (i.e., consultants, clients, research firms, etc.).
- Sellers are those who have access to respondents and can provide this access for a cost (i.e., data collectors, panel companies, call centers, etc.).
For the most part, these two parties get along and negotiate to get what they need. There has been a fair amount of commoditization in recent years, driven largely by consolidation, recession, globalization, offshoring, proliferation of DIY data collection tools and increased competition from tech companies (Google, etc.) getting into market research and pressure to contain costs. Some companies are vertically integrated and have their own in-house panel. This model is popular too because it contains costs and builds in efficiencies.
As a respondent in 2013, taking part in marketing research works something like this:
- I can join online panels and local focus group facility databases. By providing my contact information and filling out a list of things about myself such as the car I drive, number of kids, health conditions, etc., I can be invited to face-to-face or online studies and be paid for my opinions.
- Participation is voluntary and I can opt out any time. There is no limit to the number of panels, MROCs or focus group databases I belong to. I'm occasionally screened for participating in too many studies or being what the industry calls a professional respondent.
- By default, I am a person who has agreed to participate in market research studies. I'm mostly driven by the money but I also like getting out of the house now and then and I enjoy meeting people and giving my opinion.
- There are very few controls in place to determine whether I'm telling the truth to qualify for studies and not participating in too many studies.
What's wrong with the current way we manage respondents? Nothing much, unless you begin to worry how representative people who belong to marketing research panels really are - and how frequently they participate.
My company uses panels and databases just like any other company (and they do a good job) but I sometimes have this nagging voice in the back of my head (along with all the other voices!) that wonders if people who were not part of a panel would respond differently. And then you sometimes have to wonder how truthful they are. Most respondents know there's a piece of cheese waiting for them after they complete a series of tasks. People have been known to do a lot for cheese.
Changing how research gets done
This is where Graph Search could play an important role in changing how research gets done.
To illustrate this, let's go back to our dog food example. Let's imagine for a moment that Brand X is our client and they have asked us to recruit four focus groups of 10 people who are Brand X customers in Tampa.
Graph Search tells us that this should not be a problem, as we've identified over 100 people who meet our criteria. And, because this is Facebook, we actually know a lot about them. This also solves some of our screening problems. Remember that this is all self-reported data but - and this is the beauty - there is no ulterior motive. People are reporting all of this information about themselves without trying to qualify for a study.
We know that Joyce V. is an entrepreneur who works from home. She lives in Tampa and is originally from Indiana. She likes dogs, Nylabone (a dog product) and she Likes a page on Facebook called "I Love My Dog." It also looks like she went to college. If Joyce V. has an open Facebook page, we can probably learn a lot more about her. Remember, there's nothing inherently sinister about this since all of this information has been voluntarily provided by the Facebook member/respondent. And Facebook's Graph Search is providing us access to all these profiles that were previously not searchable.
Could get interesting
So here is where it could get interesting. Specifically, how do we turn Joyce V. into a focus group or survey respondent?
The options could be as follows:
- We message her using Facebook's built-in messaging system and contact her on behalf of our company to let her know we are conducting a marketing research study with dog owners. We use the standard recruiting protocol to let her know there are no sales involved and that, if she qualifies for the study, she will be compensated.
- For a quant study, we could collect e-mail addresses for the Facebook profiles that appear to qualify or are in the ballpark. We could send out an e-mail blast/invite to all prequalified respondents, again inviting them to participate in our study, in exchange for a reasonable incentive.
Of course, this raises some privacy concerns. Some people are bound to want to know how we were able to find them. This is why we need to provide respondents with the option of opting out.
Another issue is that people are not used to being contacted by companies via Facebook. As of now, Facebook is a passive system in that we can Like brands and interact with them by posting on their Timeline but there is no one-on-one relationship.
However, that could change. And as consumers become more comfortable interacting with brands in different ways on Facebook, they should also become accustomed to being contacted by marketing researchers. Just think of the transition from landlines to cell phones. People used to freak out when market researchers would call them on their cell phone because these calls would eat up their allowance of minutes. Clearly, this is no longer a barrier.
Delicately, ethically and professionally
The bottom line is that if researchers want to use Graph Search for studies, they need to do so delicately, ethically and professionally. The same ethical standards that apply to contacting respondents via other channels apply here too. Nonetheless, the promise of an incentive goes a long way. We're the only industry that finds people and then pays them for their opinions. This is an advantage we ought to leverage. Human nature dictates that people are less concerned about being contacted when you're actually looking at paying them vs. selling them something.
And then there's Facebook itself. The company has been repeatedly under fire by privacy advocates and has gotten into hot water for breaches in this area. But most of the concerns have been about the (potential) access advertisers and other data miners have to member data. The marketing research community has always had carte blanche access to consumers under the familiar guise of, "We're not selling anything. We're merely looking for your opinions and you will be paid in return for completing the study." If this has worked in real life for decades, why can't it work on Facebook? What's the worst that could happen? You get unfriended. Happens to the best of us.
Sitting on a gold mine
Now, if you've been following along carefully, you have probably figured something out else by now: Facebook is essentially sitting on a gold mine - the world's largest consumer panel of one billion members, consisting entirely of people who have shared billions of terabytes of data about their lives, their families, their jobs, their vacations and every possible minute detail of their daily existence. With Graph Search, Facebook just gave us all of us free access with powerful search features.
As a market researcher, I cannot help seeing Facebook as the world's largest MROC - one billion members and counting. There has never been a better time in history to be employed in marketing science.
But as much as I like accessing the information for free, if I were Mark Zuckerberg, I'd try to think of a way to monetize this. Helpful and invaluable as they are, the large panel providers don't have nearly the information on their members as Facebook has on its. This gives Facebook an enormous competitive advantage that could quickly (immediately!) make it a powerful player in the data-mining and marketing research business.
Not a stretch
Facebook getting into the research business is not a stretch. It actually makes a lot more sense than Google doing it. Google Consumer Surveys is a neat idea and it's up and running. Using the power of Google to find consumers is certainly nothing to dismiss. But it's Facebook that has all of the information that truly makes marketers salivate. And it will continue to do so.
But until Facebook decides to join the multibillion-dollar research industry, Graph Search is ours to mine for data and for respondents. The possibilities are limitless, even for secondary research or companies looking to expand into certain markets. It is literally a treasure trove of unlimited data just sitting there waiting to be discovered.
So the next time you need to find vegans in Portland, Ore., who listen to Radiohead and wear The North Face clothing and have a cat and one kid, try Facebook Graph Search. It's free.
For now.