Community Management, Social Media Case Studies, Social Media Strategy, Social CRM, Social Media Monitoring

Why Influence Mining is the Next Gold Rush

Inherently, we understand influence. It’s in our DNA. We know that a grizzly bear has a marked impact on its surroundings, and can change behavior in ways that even the fiercest badger cannot.

The tsunami of data being created, collected and parsed every second of every day now makes influence identification instantaneous, and possible from the comfort of your desk chair. You just need to know where to find it.

Last week, I interviewed three men that are helping create an all-new field of marketing and customer service – the emerging discipline of influence mining. Influence mining makes the old school, direct marketing notions of sorting consumers by purchase history seem dinosaurian by comparison.

  • Raj Kadam is CEO of Viralheat, a social media listening platform (disclosure: I am a paid customer of Viralheat).
  • TA. McCann is CEO of Gist, a provider of real-time intelligence on your network of online contacts.
  • Joe Fernandez is CEO of Klout, a provider of influence metrics that is being incorporated into a variety of social media tools (including CoTweet, a client of mine)

Jay Baer: What’s changed in the last two years or so with regard to access to influencer information?

  • Joe Fernandez: The amont of public-facing data attributed to a person’s name has exploded. We’ve gone from MySpace and a era of anonymous accounts and comments attributed to “SexyBear1984” to using our real names on Facebook, Twitter, and elsewhere as we build personal brands. We can now associate a tremendous amount of data to individuals.
  • T.A. McCann: We know not just identities, but also a second order of information about your content publishing frequency, density, and topical focus. And even another order that includes content interactions – who consumers and retweets your content? We’ve increased data availability by one or two orders of magnitude in just three years.

Jay Baer: Do you look at the availability of this data differently as a consumer, than you do as a C.E.O.?

  • Raj Kadam: The question we ask is whether the specific data is useful to marketers. If it is, we’ll include it in Viral Heat. If it’s not, then we don’t. We are aggregating vast amounts of data, and in our integration with Facebook for example, we don’t open up everything we get because some of it just isn’t relevant.
  • TA: For Gist, which is based around personal productivity, it’s about only revealing data in a contextually appropriate way. Each of us has to make decisions about our personal brands and about what we’re comfortable exposing about ourselves, and where.

Influence mining is primarily being used in two corners of the social Web – for identifying and engaging with categorical or brand authorities; and for adding information to customer service interactions. I asked the CEOs about how this is working in practice.

JB: How are companies incorporating influence metrics into their marketing?

  • JF: We’ve recently completed projects with Virgin America, Starbucks and CoverGirl where we emailed offers to customers based on their influence and Klout score within a particular topic.

(note: CoverGirl is a client of Marina Maher Communications, one of my clients. I did not work on the Klout campaign. note: good post about this program from Valeria Maltoni)

  • JF: Opt-in rates for those campaigns have been north of 70%, which is just astounding.
  • TA: People who are influential in a category have worked hard to build and cultivate their personal brands. Offers based on those categories enable the influencer to communicate again about those topics, continuing to build their following in that area.
  • JF: It’s important to note that influence – at least as measured by Klout – is derived over time based on consistency of content creation and interaction. We get emails in our customer service department from people wondering why their Klout score didn’t instantly jump 15 points when they were retweeted by Robert Scoble. To keep the system from being gamed, it doesn’t work like that.

JB: In addition to influencer outreach, how can companies engage in influence mining for customer service? Should companies treat customers differently?

  • JF: We hope that companies will treat every customer well, but the reality is that we’ve always categorized people, whether it’s how you’re dressed at the mall, to the car you drive. But now, it’s not based on assumptions, we can categorize based on real data.
  • There are 600 companies using Klout data in some way, and a big chunk of those are using it for customer service. Personally, I had a problem with an airline recently. I called and emailed, and received no response. Then I tweeted, and my issue was resolved.
  • TA: On inbound customer service, you can use Gist to look at the content history of the individual, and address them differently. If you see that they write about or comment on technology, you can solve their problem in a more appropriate way than if they aren’t comfortable with technology.
  • The more you can understand about the person on the other end of the phone, the higher the likelihood of a positive business outcome. If they are social media savvy, you can follow up via Twitter direct message. If they aren’t you follow up via email.

JB: Will these tools ever truly benefit the small business?

  • RK: Small business usage really requires filtering out noise, and we’re making a lot of progress on that. Viral Heat data includes location, wherever available. We have a lot of small business customers due to our price-point (note: as low as $9.99/month), so filtering and more and more analytics are important to our customers.
  • TA: Most companies are looking for listening tools to do two things: show trends and anomalies; and show cause and effect. That’s achievable for most small businesses.

JB: What are the big developments on the horizon for influence mining? What’s going to happen next?

  • RK: Measuring social media’s direct impact on sales.
  • TA: Each of our companies is a “smart agent” in some way. But there is going to be an explosion of other smart agents. “Who Should I Know” for example, is a relevant and important question that needs to be better answered. Combine that with location and mobile data, and you get real-time smart agents that tell you whom you should approach in a restaurant to buy a drink.
  • JF: Hyper-targeting. Billions and billions of dollars are spent on advertising. How can those dollars be spent more efficiently by matching products to interests?

What do you think about influence mining, and its applicability for your business? Customer service? Marketing? Knowing whether to respond to or ignore negative comments on a blog?