Over the past couple of weeks, I have the opportunity to work on a number of influencer analytics projects. One thing which has stood out consistently in all of my analyses is that influencers are not as powerful as we might think, necessitating greater care in choosing them.
As part of our READ framework for social media influence, we download and analyze the social media posts of influencers over the past 365 days in order to audit their effectiveness. We look at which influencers are truly moving the needle. To better illustrate this process, let’s take a look at the fashion sector.
We begin with standard engagement numbers for Likes and Comments:
Notice above, we see a power law curve. The top influencers garner the biggest numbers and then it’s a long tail of less-powerful influencers. From this data, we should be able to easily pick which influencers we want to work with, right?
Not so fast. Instagram likes are so easy to obtain, your cat could accidentally like every photo in your feed. Comments require at least some level of effort to leave. What if we looked at the ratio of comments to likes for each of these influencers?
Above, we see that Jaclyn Hill is the standout in the top of the pack. She has a comment to like ratio in the upper quartile, the top 25%, of her peers, while everyone else at the top of the list is in the lowest 25%, the lowest quartile. If we care about getting deeper, more meaningful engagement, we might eschew everyone ahead of Jaclyn.
This is a big deal, a major shift (pardon the on-brand pun) in how we look at influencers. We want to work with influencers who can generate effective results.
We also see that influencers are struggling with content shock as much as brands are. Let’s take a look:
Note: the left axis is logarithmic as the disparity between likes and comments is gigantic.
Above, we see that influencers are still building audience, the red line. They are growing, but not at extreme rates.
We see that engagement is mostly neutral; the orange line for likes is ascending a tiny amount while the blue line for comments is flat.
We see that the ratio between comments and likes, the green line, is the problem. Deep engagement vs. shallow engagement shows these influencers are not garnering as much deep engagement, and this has faded over the past year.
Still, with huge numbers, these influencers are still helping fashion brands reach the right audience, right?
Above we plot the total engagements (likes + comments) against the influencers’ audiences. The fuchsia line above shows that influencers’ ability to get audiences to engage at all is on a steady decline.
What can we take away from this analysis? Selecting influencers today and in the future will require significantly more data analysis capabilities if you want to maximize the impact of your influencer relations program. Simply finding the loudest people is no longer sufficient to ensure results. We have to find the most impactful influencers using a variety of metrics and carefully monitor their performance throughout our programs.
Christopher S. Penn
Vice President, Marketing Technology