Scientific Study: Are You Doing Social Marketing Wrong?

on under Social Media Marketing.

Most of the "studies" surrounding social media marketing are bunk. But every once in a while, there's a real diamond in the rough. A recent joint study conducted by the National University of Singapore and Nanjing University takes the cake for me.

The researchers collaborated with a small Asian fashion retailer to measure the impact of their Facebook page on individual consumers. They combined data collected from the Facebook API and the Facebook Data Science Team with information on over 14,000 customers, including sales data, directly from the retailer.

Now those are some metrics.

Next, they categorized posts and comments into these groups:

  • User-generated content - the post or comment was made by a user, not by the retailer
  • Marketer-generated content - the post or comment was made by the retailer
  • Information Valance - The communications were analyzed with advanced text-mining tools and sentiment classification models to determine whether the message was positive or negative
  • Information richness - The text-mining tools were also used to estimate the "number of concepts" extracted from each message.
  • Directed communication - The message was communicated directly to the user
  • Undirected communication - The message was communicated broadly

The study was also careful to avoid hasty correlations. They controlled for confounding variables like:

  • Pricing
  • Individual consumer posting volume
  • Peer social networks
  • Time levels
  • Promotional events from the retailer's marketing calendar
  • Age
  • Income
  • Gender
  • Various self-selection issues

In short, this was serious science, not some amateurs looking for exposure with a little case study.

Here are some of the results, and remember, these are after controlling for the factors listed above:

  • Joining the Facebook page resulted in a minimum average increase in lifetime value of over $22. This effect was explained entirely by the interactions that took place on the page, and not by the simple act of joining the page.
  • Overall, user generated content had a much stronger impact than marketer generated content.
  • Information rich communications with other users, whether direct or indirect, and even positive or negative, all had a positive influence on sales. Paradoxically, the most effective communications were positive, indirect communications from other users. Perhaps even more strange, during direct communications with other users, positivity or negativity didn't seem to impact sales. Again, information richness always improved sales in these user-user communications.
  • Things were quite different with marketer communications. In this case, only direct communications impacted sales. The information richness of those communications didn't matter. Only the positivity of the message mattered.
  • Crucially, information richness from direct communications with other users transformed their products into price inelastic commodities. (The elasticity of demand associated with these communications was 0.006, about as low as things can get.) In other words, they could, in theory, charge those particular customers much higher prices with very little impact on their demand.

There are some seriously important lessons here, and pretty much everybody in social media marketing is getting at least one of them wrong:

  • Trying to bury negative reviews, or limiting social capabilities in order to prevent them, is pointless. Information rich comments from other users, even if they are negative, have a positive influence on sales.
  • Your role, at least on Facebook, is fairly straightforward. Communicate directly with consumers in a positive way, and do what it takes to create an active community that produces plenty of user generated content.
  • Your primary goal as a community leader is to get users to share as much information rich content as possible.
  • While indirect user sharing is more effective when the content is positive, it is still effective when it is negative, and positivity or negativity doesn't matter when users are interacting directly with each other. If you can encourage positive communications, do it, but don't force it to the point of reducing user activity.
  • Getting page or post "likes" isn't going to do anything for you. You need an audience that interacts with itself (or you need to interact directly with every single one of them) in order to have a positive impact on sales.
  • Encouraging direct communications can transform your product from price elastic to price inelastic (within a reasonable range) among the users involved in those communications.

Of course, one confounding variable I always question in any of these studies is the issue of whether the marketers themselves could have done better. Was the user-generated content more effective because it was from users, or was it only more effective because the marketer wasn't sharing compelling content?

This study can't answer that question, but that doesn't mean it's not worth asking.

So there you have it: scientific evidence that social media can and does impact sales (if you look waaaay past "likes"), and some data-driven tips to actually make that happen.

It's about damn time.

Image credit: Simon