Image: "Station Bullhorn," by Nate Beaty
My thinking on influence continues to evolve, in no small part due to some excellent exchanges I’ve recently had with my own group of “influencers.” Specifically, I recently attended a session on online influence at Blogworld given by my friends Chuck Hemann and Matt Ridings. The session was fascinating because Chuck and Matt come at the problem from two different directions – and the good news for marketers is that those directions are not mutually exclusive.
Where Matt spun me around was with this subtle point – we are consumed with focusing on influencers, but ignoring the influenced (and you should download his excellent presentation here.) As Matt noted, finding people at the moment they are actually being influenced online is really about identifying a future customer at the point of need, and turning them into an *advocate.*
Who Are The ‘Influencers’?
One of my “primary influencers,” Tamsen McMahon (who was also at this session) is fond of noting that “we are all influencers.” We all hold sway over some group of people, about some topic, in some context. Yet, the current crop of automatic influence measures will never do a *great* job at picking up on this subtlety, because they are engineered to look at aggregated data and predict a specific response. In essence, they are built to look at an abstraction – my online “activity” – and extrapolate a predictive measure: how well or poorly I am likelyto disseminate a given message, or encourage a given action, based upon my previous behavior.
There is merit to predictive measures, and I’m not going to bash them here – they are getting better. But it strikes me, taking a cue from Matt, that we are looking at abstractions and “hoping” for a specific, when we could just as easily (more easily, in fact) look for specifics and then reward future, similar behaviors. In my day job, we know that one of the best predictors of future voting behavior iswhether or not you voted last time. Influence works in a similar fashion – we need only flip the funnel around. Rather than (or in addition to) attempting to predict specific future behavior from a mass of general data, we could also look for specific instances of *actual* influence being wielded in situ, and then reward/encourage future behaviors accordingly. To go back to my voting analogy, the best way to determine whether or not I am an influencer about, say, cars is to observe me actually influencing someone to test drive or consider a given car, and not merely by the fact that I talk about cars and generate retweets.
The ‘Influenced’
I’ll give you a specific example that occurred online last week between Matt Ridings and I (and Matt, to his endless credit, worked it directly into his Blogworld talk. Have I mentioned yet that you should follow him on the Twitter?) Last week, I observed Matt on Twitter asking for a recommendation about comfortable earphones. We are close friends, so he knows that I am a bit of an audiophile (and by “a bit” I mean I have a problem.) I replied to him that he should try equipping his earphones, whatever the model, with foam tips from a company called Comply – I use their tips on all of my in-ear ‘phones, because they are super comfortable and preserve the sound quality.
The attentive folks at Comply were clearly monitoring these conversations, and they helpfully popped in to hook Matt up with some to try. Comply was listening at the point of need, and they acted to provide a sample to a potential customer. As Matt points out, however, there is a more subtle point to be made with this first level of interaction – by listening to Matt at his point of need, and solving his problem (hopefully with a great product), they might have just created something beyond an “influencer,” they might have just kindled the spark of anadvocate. Matt may or may not be a topical influencer about audio, but if he has a great experience with Comply, he will certainly be a compelling online advocate for their brand.
Rewarding Matt at the time of need, however, is just the first level of this model of influence. Matt is dead on that focusing on the influenced is an incredibly efficient way of building advocacy. None of the online influence measures would have tagged me as influential about audio – I simply don’t talk about them online that much. But my friends know that I have a wealth of knowledge about audio, and I’ve recommended countless products over the years. My aggregate behavior would never predict this specific interchange; yet, there it is. A few tweets online, and a product is sold. Does Klout think I am influential about headphones? Nope. Am I? Demonstrably.
Finding this specific interactions – the “have you actually voted in the past” of online influence – is an easy and efficient way to identify influencers. This isn’t about predicting whether or not I might wield influence about headphones, it’s observing directly that I was the catalyst for an action that the brand cared about. Comply doesn’t need to look at my aggregate profile and “hope” I’ll talk about their specific product, because now they know that I can and will actuallymove their product. I’m an advocate.
Cultivating The Advocate
As Tamsen pointed out in our Blogworld exchange, however, rewarding Matt and responding to that exchange was just the surface level. Tamsen (who, besides wielding considerable “clout” with me, is also the VP of Digital Strategy at Allen & Garritsen and a supah smaht cookie) noted that the second level of engagement for Comply would have been to reward me for my advocacy – or at least to have provided some form of incentive to encourage future recommendations. Had Comply simply butted in and pushed their productwithout my recommendation, it would have been seen by Matt as intrusive. But my advocacy warmed up the prospect, so to speak, so the “intrusion” was welcomed. Finding an actual, brand-specific influencer for Comply should begold to them, and they should do more to cultivate those specific relationships where they are discovered.
Purists might point out that this kind of “incentivized” advocacy could be seen as my simply being bought and sold asking an affiliate marketer. But, that’s really my problem, isn’t it? After all – in this model – the proof is in the pudding. I either encourage trial or I don’t, right? If I am no longer seen as influential about this brand, it wouldn’t be be because I am compensated, per se. It would be because I stopped being a compelling advocate and encouraging online interactions similar to the one I had with Matt. That would only happen if somehow my friends and online connections questioned my integrity. That’s kind of a “life” problem that I don’t anticipate having.
The Story
Still, I’m not angling for free Comply products. No, my reward is the same as anyone else’s who genuinely and passionately recommends a product to a friend: that they tried the product and loved it, thereby making the recommender (me) feel smart and loved. Right? So the “reward” I really want is for Matt to hop back on the Twitters and tell the world that he tried a product based upon my recommendation, and it was awesome – making me, by the transitive property of awesome, also awesome. This would make me happy – but (and this is thethird level of influencer engagement here) it would also make Comply happy. I’ve shared my story with Matt, so the next step is for Comply to help Matt “complete the sentence” by providing an incentive to (publicly) thank me for my awesome recommendation. Products aren’t sold in social media on features – they are sold on stories.
Sharing those stories online brings in the silent audience – the people lurking, following me or Matt – who quietly file this anecdote away until they reach theirpoint of need. It’s a simple matter for a company to add “how did you find out about us” to their order page (and, by the way, to reward customers somehow for sharing that data) or to do some other form of primary market research. Doing so might allow Comply to see that my little online exchange with Matt might have actually led to 25 or 30 sales. Who knows. And encouraging Matt to close the loop on this “story” more than doubles the chance that future, prospective customers will observe this advocacy.
The Shift
Once that “story” has been completed, there is a lasting social record of my ability to recommend a brand in a specific category, and instigate an action that brand cares about (i.e., more than a “retweet.”) This interaction is public, and valuable to Comply. It’s more than that, however, and that is where a shift in perspective is required to access the fourth level of this particular model. Comply operates in a category with partners (earphone manufacturers) and competitors (Monster, for one). If I were the maker of a high-end earphone (say,Sennheiser), I would be doing more than simply monitoring social media for mentions of my brand, or even my competition – I’d be searching specifically for interactions just like the one that Matt had (and setting up that kind of smarter research station is the kind of work I *love* to do.) Again, no aggregate, predictive measure would discern that I am an influencer about headphones, but Sennheiser could easily see from the public record of my dealings with Comply that, actually, I am.
Sennheiser could then do the work to determine whether or not I would be a good potential “influencer” or even advocate for their brand as well. And that is a straight-up influencer campaign. Allow me to sample the product, blow me away (that’s super important) and then cultivate a relationship. If the product is great, the interaction positive and a relationship maintained, Sennheiser stands a decent chance of being in my consideration set the next time someone asks me to recommend some headphones. The only difference here is that they didn’t approach it from the fat end of the funnel by using a predictive tool to find people online who are active and talk about headphones – they used a diagnostic tool to find someone who can get someone else to try a headphone or headphone related product.
Beware False Choices
Again, predictive tools have their place, and I am suggesting a different way, not a replacement way. And the top of the funnel methods (using automated influence lists) are really better suited for awareness, while the approach described in this post is more about trial and conversion. So I’m not setting the two approaches in conflict, and neither should you. And in both cases, you should check your work. If you are interested in a pretty useful, do-it-yourself way to approach measuring your efforts, I wrote a post on a true measure of influence a few weeks ago that I think you’ll find useful.
I’m not in the business of running influencer campaigns (and Matt Ridings has forgotten more about that then I’ll ever know) so I’ll leave the details to you. I am, however, in the business of asking better questions for businesses in order to get better results. Tying research – doing your own work – to these efforts is crucial in the development of a learning organization, and it is the learning organization that succeeds while the others use the same tools and information that everybody else has. Ask better questions, make smarter decisions – but above all, check your work. That is the only way to lift the “influencer” campaign from a mere tactical interaction, to one that actually might just alter how you do business and affect the theory of the firm. Darwin would approve.
Article from http://brandsavant.com/flipping-the-funnel-of-influence/
http://SocialBusinessToday.net -
The Best in Social Business
To use the Foursquare sync and get the $25 deal, just use the app tocheck in to a business, then tap the button that says "load to card." Shortly after you pay, you'll receive notification that you just got at $25 American Express account credit, which will appear on your statement around five business days later.
![Crowd Power: What Is Crowdfunding? [INFOGRAPHIC]](https://lh3.googleusercontent.com/blogger_img_proxy/AEn0k_uN-_dlm3BjfBNo6cGM0rV6A8dB5Aq6ZF3n47Y7_4yhPwqMe2bLU_GKMzfeVkdvC7wu_5ctHf3XjXo7-LxAdvRgCb3hoFUEKBWg7IS94hY_na4sjysvbR7DSNGTOZcIZo_0DjnGdVdRFK-pWAe9AJwkeQ_nmBXjCpHdubd4GCqRCV8TcNpsvX1H5F4F=s0-d)
