Tuesday, March 18, 2008

Utilizing Social Networks for Market Research: Friend or Foe

As social networks grow it is inevitable that market researchers will try to and tap this market (mainly consisting of teenagers and young adults born between 1980 and 1997). A new company, Peanut Labs, is working to capitalize on this opportunity by creating an online panel from social network members.

So what makes Peanut Labs’ panel different from other online panels? All respondents are recruited directly from multiple social networking sites using non-monetary rewards. According to Peanut Labs website, there are 2.8 million respondents in their panel across 72 social networks and online user reach is up to 12 million.

With huge concerns around professional survey takers and panel overlap in online research (as noted in this Quirks Article), the idea of having non-monetary rewards is a big selling point. Instead of money, respondents are paid in virtual currency according to their social network. For example respondents on Facebook can earn a currency called “munny” for their virtual pets at (fluff)friends. Respondents earn 100 “munny” for their first survey, 200 “munny” for each new survey they complete, and 40 “munny” for being pre-screened out of a survey ($1 USD = 200 “munny”). This method is considerably cheaper than paying respondents actual money and marketers can therefore offer higher incentives to not only respondents that finish survey but also to those who do not qualify. The logic behind this is people will make “munny” regardless of qualifying for the survey, and will consequently be more likely to answer questions honestly.

At first, I sit back and think “Wow, this is pretty amazing.” Most of our concerns as online quantitative researchers could be solved (or at least this is a good start). Then I begin to think about my own personal social networking habits. I am currently a member of Facebook, Myspace, Ringo and LinkedIn. Most of my friends and colleagues are also members of multiple social networking sites. Some are also members of panels such as E-Rewards and Greenfield Online. In addition to being members of multiple sites, we also use multiple email addresses when logging into these sites. This makes it considerably easier for respondents to misrepresent themselves and brings up concerns on the issue of respondent duplication.

Take this scenario for example: A respondent signs up to take surveys through multiple social network sites by using different email addresses on each site, how will a client be able to have confidence in the sample and know that a respondent did not take the same survey three times on three different social networks (especially if the target for a survey is a niche market)?

I believe utilizing social networks and offering non-monetary incentives is a great step toward eliminating some online research woes, but until respondents can be verified and duplication is eliminated across panels, social networks, etc. can anyone really say that one is more reliable than the other? What are your thoughts?

5 comments:

Anonymous said...

Hi Kris564,



Thanks for taking note of Peanut Labs. Your analysis of virtual incentives is absolutely right. Because they are completely relevant to the users’ online community, they find the surveys as relevant “fun activities”, and are encouraged to participate in the surveys. You also mentioned the industry concern with online data quality, and in fact, we actually just launched a new technology this week called OptimusID that filters out professional survey takers and other suspect respondents on online panels. It is currently in beta, and is being tested by more than 20 research firms from the Honomichl. We hope that this technology will help bring the industry closer to achieving high quality data when conducting online research.



-Ali

finnstones said...

Loyal reader of your blog. Do you have a twitter account...if you do, will you share it with a stranger?? Twitter should record your greatness.

- overboard fanboy

Tom H C Anderson said...

Not sure if giving Mummy points solves these research problems. But I applaud the effort and application of any Next Gen Research methods. In terms of Facebook and Linkedin, the only two SNS that currently really seem to matter, Facebook is already offering polling feature, and Linkedin is soon to follow. Who knows perhaps someday they will both be in competition with PeanutLabs, Greenfield, SSI and all other sample providers? The new Linkedin polling initiative is discussed briefly here: http://www.tomhcanderson.com/2008/05/04/facebook-and-now-linked-enter-the-%e2%80%9cmarket-research%e2%80%9d-game/

Anonymous said...

Thanks for the interesting post.

I too have concerns regarding social networks and market research - knowing who your participants are is vital to the integratity of the data.

Bespoke online communities set up facebook-style but with participants verified can provide clients with engaged participants, organic conversations and rich insights.

Anonymous said...

Thanks for the interesting post.

I too have concerns regarding social networks and market research - knowing who your participants are is vital to the integratity of the data.

Bespoke online communities set up facebook-style but with participants verified can provide clients with engaged participants, organic conversations and rich insights.