Is Influencer Theory Garbage?

Duncan J Watts recently published a study debunking the role of influencers in synthetic virtual worlds, then debunking them again by showing how ads spread through the online world. I think the problems with the thesis are

  • machines do not have emotions
  • most of the people spreading the ideas in his studies did not have a large potential gain as a potential outcome of sharing the idea
  • we live in a time of such an abundance of information that information / knowledge workers need to trust filters
  • some filters have publicly available stats showing 10,000's of people trust and follow them
  • the easiest way to target influencers is not through ads, but through content which gets exposure in some of those trusted filters...if your idea starts in an ad box you already tuned out many influencers

The Fast Company article covering the research goes on to state the following

As Watts points out, viral thinkers analyze trends after they've broken out. "They start with an existing trend, like Hush Puppies, and they go backward until they've identified the people who did it first, and then they go, 'Okay, these are the Influentials!'" But who's to say those aren't just Watts's accidental Influentials, random smokers who walked, unwittingly, into a dry forest?

In some cases that might be true, but online you can learn communities and individuals well enough to create content targeted around their needs / wants / passions / biases / identities. You can predict the viability future ideas with some degree of accuracy. And there is so much data to study that virtually anyone can pick up the patterns and start spreading ideas within a few months.

I have changed a few words in a blog post to change the angle of it to target certain individuals. My success rate with getting a mention from the specific personality or person I was targeting is much too high to be an anomaly. Not only have I taken past ideas from other markets and applied it to my market, but I have taken some fundamental social and psychological principals and value related ideas, applied them to markets I know almost nothing about (and have no influence in), and still over half of those linkbait ideas go viral.

And once you give an idea exposure on a leading channel or two (not as an ad but as editorial) you display social proof of value and start the cumulative advantage process.

I am not trying to toot my own horn. Just trying to write from experience rather than theory. Some people, like Andy Hagans, are way better at launching linkbait than I am. The reason that linkbait is so powerful is that it can be so targeted, and it targets the foundation of the web's value system, which typically does not look like an ad - the link.

Published: January 31, 2008 by Aaron Wall in marketing


January 31, 2008 - 8:05am

Just a comment: I see two types of influencers on Internet: Those who know a lot about SEO and marketing and they make linkbait intentionally, and those (the true influencers) who simply generate content with no linkbaiting objectives, but the content is so excellent that we feel the necessity of linking their content. However, some few people belong to the two types.

January 31, 2008 - 1:03pm

Well, there have been a string of reports in the mass media about researchers in social networks who are challenging the big results of the 1970's. That's how science works.

As an SEO/SMO guy, I've been moving towards a link building strategy based on getting referrer trafic rather than getting "page rank". I can get real-time information about the # of inbounds I get from links that I made yesterday, whereas I can only guess what's moving my rankings on Google.

One thing I'm thinking about is: should I be trying to get a few high-traffic links or lots of low-traffic links? Most sites get a disproportionate amount of traffic from a few inlinks: for instance, the top referring site has sent one of my sites 150 visitors this morning. (Some days it sends me 100 times more!) Other than search engines, the next site sent me five, but I've got about ten or twenty that are that big.

Looking at referrer data from a big sample (>> 10k sites), I've seen that most sites show the same pattern. In real life, it does seem important to get a few big links.

January 31, 2008 - 2:55pm

I think if you meet the criteria to get the few big links that it has a cascading effect where you get many follow on smaller links.

January 31, 2008 - 5:13pm

Yes, Aaron, it's certainly true that links multiply and that big links multiply well.

In planning link building activities, I think about this constant that I call "k". "k" is the ratio of links that I actually get compared to the number of links that I try to get. (A "try" could be sending an e-mail, or it could be posting a comment to a blog, or any action that might produce a link.) To make things a little more definite, let's say that we're counting the number of links one week after the "try".

It's hard to build links for certain kinds of sites: some people report k ~ 0.01. I've been fortunate enough to work on some project where k=2 or 3.

Human-powered viral spread is one reason for overunity k. Another one is automated spread. I like getting Wikipedia links, for instance, because people are always making copies of wikipedia, sometimes with nofollow turned off. If you have some links in wikipedia, you'll find that you keep getting more links all the time, which is good for your link velocity. Also, I've found that scrapers tend to hang out around certain high-quality RSS feeds, so sometimes a "good" link will turn into 30 links. Maybe they're not the best quality links, but they are links.

Add new comment

(If you're a human, don't change the following field)
Your first name.
(If you're a human, don't change the following field)
Your first name.
(If you're a human, don't change the following field)
Your first name.