Market-research consultancy Anderson Analytics released findings from a study this week that emphasizes the wealth of consumer information available on social-networking sites (SNS) -- information that goes far beyond the explicit data listed in each consumer's profile. The study relies on information derived from the social-networking site LinkedIn, and included information from the records of more than 53,000 individuals, filtered and crunched using Clementine, the data- and text-mining solution from analytics pioneer SPSS.
In addition to directly surveying 793 LinkedIn members, Anderson Analytics looked at 2,000 "seed" profiles and the first-level connections of those 2,000 profiles.
The objectives of the study, explains Tom Anderson, founder and managing partner of Anderson Analytics, was twofold:
- To show how text mining and data mining can be used to leverage data from social-networking sites; and
- to show how to gain access to a "more-focused population of respondents" contained within a social-networking environment, rather than seeking participants from unknown sources on the Web through companies that specialize in mass-market polling and surveying online.
"The use of SNS is also more economical [and] easier, and an efficient way to reach millions of consumers who could never be found in focus groups or research panels," Anderson said in the release.
According to the study, LinkedIn members were asked to complete a 10-minute online survey. Of those who responded, only 14 percent reported prior participation in online survey panels. Moreover, the analytics indicated that those who had no prior experience as panelists, or perhaps had left a panel, tended to have a higher income and stronger purchasing influence and budget. "This seems to indicate that, once you get promoted, you're less likely to participate in these online panels," Anderson says. Ironically, 66 percent of this survey's respondents fell into that category -- a reflection, perhaps, of the highly professional nature of the LinkedIn demographic. The study warns marketers that there may be a risk in relying solely on an online panel, and suggests that marketers need to tap into high-quality, self-defining networks if their intent is to reach a particular target group.
"Marketers...know that social networking is a new phenomenon that's very powerful and everyone's trying to understand how to leverage it, but marketers haven't really gotten their heads around it yet," Anderson says. With LinkedIn in particular, the challenge lies in the mining of the varying levels of connections. (The site links members to their first-, second-, and third-level connections.)
Anderson Analytics is also attempting to track the relationships of "Gen X to Z" through Facebook in a similar fashion.
The challenge with surveying a random "representative" sample, Anderson says, is that, ultimately, you don't know who each person is. On the other hand, today's technology allows companies to identify the 10-digit machine ID stored on each computer user's system board. That information uniquely identifies the machine itself, taking identification beyond the more commonly used Internet Protocol (IP) address, which reflects only the connection to the Internet rather than the person making the connection.
The use of this newer technology attempts to control for discrepancies, such as multiple survey entries, but Anderson says the combination of the technology with social-networking data can be powerful. "You still can't tell if that's the mom in Ohio or a Fortune 500 [chief marketing officer], but with LinkedIn, if that person has 100 connections in their industry and 5 referrals, you know that person is [most likely] who they say they are."
Using data- and text-mining, Anderson Analytics gleaned information from several unorthodox data fields on LinkedIn, such as occupation headlines (as opposed to, say, "Job Title"). Based on behavioral and attitudinal data such as number of connections, logins, usage style, and networking interests, the data-analysis system categorized individuals into four groups. Anderson Analytics identified the four segments as follows:
Savvy Networkers (9 million):
- early SNS adopters (including sites such as Facebook);
- more tech-savvy;
- more likely to use LinkedIn for purposes other than mere job-searching;
- have an average of 61 connections, the highest of any of the four groups;
- usually have "Consultant" in their job description; and
- have the second-highest personal income ($93,500).
Senior Executives (8.4 million):
- less tech-savvy;
- using LinkedIn to connect to their existing corporate networks;
- have "power" jobs;
- likely to have been invited to join LinkedIn by a colleague;
- have an average of 32 connections;
- have titles such as Owner, Partner, Executive, or Associate; and
- have the highest average personal income ($104,000).
Late Adopters (6.6 million):
- less tech-savvy;
- likely to have succumb to numerous requests to join;
- tend to connect only to close friends and colleagues;
- have the fewest number of connections, with 23 on average;
- have titles such as Teacher, Medical Professional, Lawyer, or the word "Account" or "Assistant" in their job description; and
- have a personal income of $88,000.
Exploring Options (6.1 million):
- fairly tech-savvy;
- use SNS for corporate and personal interests;
- have an average of 34 connections -- the second-highest total;
- may be employed, but are open and looking for other job options; and
- have the lowest personal income ($87,500).
Using mining solutions, Anderson says, "we can obviously predict which of these four segments [a given person] would fall into." Armed with that level of insight, marketers can accurately target their desired group with advertisements, or surveys could be designed to solicit responses from a sample based on specific characteristics derived directly from online profiles.
Anderson says that one current client is interested in speaking to the vice president of human resources at Fortune 500 and Fortune 1000 companies. The ability to target that specific sample, he says, "wouldn't be possible anywhere else, obviously, except through LinkedIn. Nobody [else] has those connections."
LinkedIn itself clearly realizes the potential goldmine that has developed on its watch: The company announced in late October the launch of LinkedIn Surveys, a service allowing market-research professionals to conduct B2B primary research among its members.
"Marketers need to understand how powerful social networking is -- and leverage it," Anderson says. "The tools are available to us now."
News relevant to the customer relationship management industry is posted several times a day on destinationCRM.com, in addition to the news section Insight that appears every month in the pages of CRM magazine. You may leave a public comment regarding this article by clicking on "Comments" at the top; to contact the editors, please email editor@destinationCRM.com.