Under(estimating) The Influence.
If there's one thing I distrust more than the reams of unthinking nonsense written about influencer marketing, it's the idea that self-reported attitudinal research has any validity.
I've seen it claimed that the graphic above shows that social media "influencers" have little effect on people. I treat that claim with the same credence that I give to people who claim they're not influenced by advertising.
But, equally, I wish that the influencer marketing zealots would read a little network theory and accept that virality is dependent on the structure of the network rather than the popularity of the "influencer". Receptivity trumps volume every time.