Predicting where the social dice will fall


When digital shop Organic hired economist Jason Harper to evaluate ad campaigns, Harper had a brilliant idea — rather than use social media to push sales messages or monitor fuzzy “sentiment,” how about crunching social media mentions to predict whether current ad campaigns will hit their lead targets? In other words, every campaign that penetrates consumer word of mouth tools such as Twitter’s tweets or Facebook’s Likes has an ebb and flow — a baseline of buzz before the campaign, a spike in mentions as the campaign message scales, and then a radical falloff of chatter. By measuring the velocity and acceleration of the curves, Harper came up with predictive models showing whether in-market advertising was on track.

MIT’s Technology Review reports Harper has successfully done this for auto companies and brands such as Kotex. The article mentions only a few case studies, and one client, Chrysler, eventually moved on to another agency, but we love the concept. Social media may not be about pushing sales or pulling mentions; instead, it could be about monitoring what consumers are saying about your ads in traditional channels, and then adjusting your course accordingly.

Image: Rob W.

3 thoughts on “Predicting where the social dice will fall

  1. Ben, while I agree with this approach, don’t you also think that by tying in proximity based metrics via mobile also have a role here? The way a brand can now engage someone at retail or at an event in real time opens up many opportunities to quickly test various content to see which works and modify in real time. Its similar to a traditional DM strategy where you would have a control piece and several others to test and learn against to tweak a message. I also believe that the various tools to measure sentiment play a still factor here and are just another component within the predictive tool box.
    As always, another great pov. Thanks.

  2. Michael, I love the idea of proximity-based metrics and incentives … and am amazed how difficult this remains. Marketers still seem stymied by an ability to influence people near the point of purchase; some game-mechanic type systems such as Foursquare are trying, SMS response and QR Code snaps are also cumbersome attempts, but the verdict is still out.

    Perhaps devices have to get better before marketers can manipulate the message through them near the point of purchase. LBS systems seem hung up in game play to get early adopters, and I’m wondering if that is just a fad. Until the masses adopt forms of LBS as part of their shopping modes, influence there, and measuring it, will remain difficult.

    I’m open to updates, lmk if you know agencies, brands, or channels that are cracking the proximity code.

  3. Thanks Ben. The channels and platforms you mention, are certainly components to cracking it. However, It’s about using these platforms within a broader communications program that positions LBS as a utility, instead of a game. By providing ongoing value through the engagement and offering some aspect of reward, you’ll be able to more relevant and capture better metrics. It’s an area that I’m deeply involved in right now, and as I uncover more I’ll be sure to share.

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *