Duncan Watts first came to the marketing world’s attention in 2003, when a New York Times article positioned the Columbia University professor as the anti-Malcolm Gladwell. While Gladwell’s 2000 book The Tipping Point laid out an entertaining theory that fads like Hush Puppy shoes and the book Divine Secrets of the Ya-Ya Sisterhood spread because of groups of people called influencers, Watts has argued that that model is deeply flawed. Watts, now a research scientist with Yahoo, charges that influencers have been poorly defined and that scientific data show you’re just as likely to spread a message or product by targeting a random group of consumers as you would by going after so-called influencers. Mostly, Watts says marketers should be much more scientific in their approach, especially as social media grows in importance. Excerpts from Watts’ conversation with Brandweek are below:
Brandweek: What’s wrong with the influencer model?
Duncan Watts: The claim that influencer matter or are important or influencers drive brand awareness, when you scrutinize them carefully, they turn out to not really be very meaningful. Or to put it another way, everyone thinks they know what an influencer is and everyone thinks they know why they matter, but everybody thinks something different. Is an influencer the hipsters in the East Village or Oprah Winfrey? What makes Oprah influential is very different from what makes the hipster in the East Village influential. And so by failing to differentiate carefully between all these different types of influencers you really undermine the ability of the theory to say anything predictive.
What about a phenomenon of this decade like Crocs. How did that catch on? How do things like that spread?
I don’t think anyone knows. Somebody asked the publisher of the surprise bestseller Eats, Shoots & Leaves why that book was so successful and he said it was successful because lots of people bought it. Maybe that’s an honest appraisal. Hits are highly unpredictable. It’s very difficult to even retrospectively go back and show that there are certain kinds of consistent attributes that result in being popular. So Crocs is a good example. Nobody would have seen that they were going to be such a hit, so when confronted with that sort of evidence, people want to say ‘OK, maybe you can’t predict what attributes make something popular, but the reason why they’re popular is certain special individuals promoted them so now you can predict.’ It’s all about this desire to make predictions and to make them in terms of simple intuitive models.
So what would your advice be to a marketer trying to learn from things like that?
My first advice would be stop fooling yourself. If you have the wrong model of the world, getting the right model of the world first requires acknowledging that you have the wrong model. Marketers have been chasing influencers for a decade and they haven’t found them. And the reason is not that influence doesn’t matter. It may very well matter. The reason is not that some people are not more influential than others — that may also be true.
The reason is that history is a very poor guide to the future. Just because the hipsters in the East Village were wearing Hush Puppies and suddenly everyone else started wearing them doesn’t mean that you can go out and get the hipsters in the East Village to wear your product and it will be popular. To put it another way: Hipsters in the East Village are wearing stuff all the time and it doesn’t always become popular.
What about these consumer packaged goods companies going after mommy bloggers? Is that a waste of effort?
It may or may not be. The question is: What’s the evidence? What I was actually trying to advocate was if you actually want to figure out how influence works and you want to intervene to exploit social influence, you need to start doing it in a more scientific manner, which means you need to do experiments and collect data. You need to stop assuming that you already know how to do it and try to figure it out. This was very unpopular. This was not a piece of advice that anyone was interested in. I went around and I met with a bunch of people in the marketing world and I said look, if you want to find influencers and you want to find out who will be better at promoting your product or generating diffusion you need to run experiments.
Let’s say you have a product to promote and you want to give away free samples and you think if you give this to the right people they will A. Like the product and B. Tell their friends and get all their friends and their friends’ friends to buy the product and I’m going to sell so many units through the diffusion process. So who should I give them to? This is where everybody has their pet theory about who’s influential. The point I was making was if you want to figure it out, you have to do experiments where you take a bunch of people who you think are influential and then you take a control group. Or alternatively you say I don’t know who’s influential, but I have a bunch of different theories. Maybe I think the popular kids in school are the influencers, but maybe I think it’s the guys in school who are really into technology or into music who are really passionate about it.
Or maybe it’s the kids who aren’t the most popular, but are in a lot of groups. So those are three different theories about whose influential. You’d like to test those groups against each other and have some kind of horserace in a controlled manner and you give away samples to people of different types and you actually measure the uptake. This is the kind of thing you could do if you were serious about finding out who the influencers were.
But marketers aren’t interested in that?
What I started to realize after several of these conversations was the whole influencer theory is actually more of a rhetorical device than a theory. It’s not like people actually have an explicit theory of who is influential and they go out there and they use that theory to decide who to target. It’s more that they do whatever they do. They throw parties or they give away free samples or they advertise in particular publications that brand themselves as reaching an influential audience and run some campaign in their normal manner. And if it works, they say, “We reached the influencers.” And if it doesn’t work, they say “We didn’t reach the influencers” or “We weren’t able to make them help us.” In that sense, it’s just a rhetorical device to help you explain the randomness that you actually experience in the world. I thought people actually believed this and in believing it they actually used it and that’s bad because people should be thinking more systematically. So I was sort of trying to say to them “OK, if you want to find the influencers, I’ll tell you how to do that.
And if it turns out that you can’t find them and in fact you’re better off seeding things randomly, for example, then that’s great because you stop wasting your time and energy. But the reaction to that was I think informative that in fact they weren’t doing that to begin with…The irony I think is that targeting ordinary people is probably what marketers are doing and if something takes off they say “Oh, we reached the influencers.” So it’s one of these impossible-to-falsify theories because who you identify as an influencer is always after the fact.
What are you working on at Yahoo?
We are still interested in social influence. What we’re trying to do now is use network data in a non-experimental manner to try to see — if you look at a diffusion of a particular product over a social network just using observational data-you can ask a question about if this product gets taken up by a group of people and subsequently their friends adopt it — it’s really like a real-life simulation of the models we were running. In a large network, you’re dropping a seed and asking how much do each of these seeds grow until they stop. Now we have a bunch of trees, diffusion trees, and some of them are very small and a few are very big. And what you’d like to be able to do is say “Can I predict the size of the diffusion trees based on what I know of the seed?” This is the influencer’s hypothesis, that there’s something special about the seed that predicts the size of the tree.
We ran that in simulation results and now what we’re trying to do is replicate that using empirical data, which is not quite the same as dropping it in the real world, but it’s the best we can do at this point. We are still pursuing this question of what predicts success. You can throw into that model both the attributes of the people and the thing itself, the thing that’s diffusing and you can throw in attributes of the network. That’s where it might get interesting. If you found for example that you can’t really predict what product is going to take off and you can’t predict which individual is going to be responsible for spreading it, but there are features of the network which are conducive or not conducive to spreading. That would really be interesting. Then you’d have a true network theory of diffusion. If you can go out and measure certain features of the social network you can actually optimize your targeting around those features. From that perspective, the influencer hypothesis is just the simplest network statistic that you can measure, like who had the most links or who had the most Twitter followers. Let’s call them the influencers. So a strategy of picking the most highly connected nodes is like the very simplest network measure that you can think of. You might say “Well, that doesn’t really work because people with high degrees of [connections] are followed by people who aren’t that interested in what they have to say.” Or the people who follow them are people who no one else is interested in reaching. So you can reach a bunch of people, but it never goes anywhere.
You might find a mix works better?
The insight that we got out of the simulation models was the network attributes that was conducive to diffusion is: Easily influenced people influencing other easily influenced people. If you get a long sequence of those individuals, you can propagate for many steps. Now the interesting question is does anything like that rule come out of the real-world data?