Key Article Takeaways:
- Best Practices in risk management (including influencer marketing) include:
- A portfolio approach
- Understanding the variability (Beta) in performance.
- Always estimate views based on medians, not averages
- Video Views are limited on the downside at zero, but there is no bound on the upside. That's where virality comes in.
- Working with multiple influencers mitigates the "all eggs in one basket" risk and increases the likelihood for a post to nail it and outperform
Whether it’s bitcoin prices going to the stratosphere (and then back down) or the gyrations of GameStop’s stock price, nothing gets our attention more than rapid motion: especially rapid motion that implies that human behavior has changed in some fundamental, unstoppable way. Our brains, which are generally not great at math, turn out to be extremely good (or bad, depending on your perspective) at extrapolating a trend and assuming it will continue forever.
So it is not particularly surprising that the birth of influencer marketing was boosted by attention grabbing headlines about content that had gone “viral” on YouTube or elsewhere. I remember a Board member forwarding me an article with eye-popping numbers about beauty vloggers and if I’m being honest, it changed the way I thought about the industry. "How fun would it be reporting results like THAT to a client??" I remember thinking.
In contrast to the casino cowboys messing around with crypto currency, influencer marketing professionals just want to be liked so very quickly everyone was putting out insightful blog posts like, “Going viral is *not* a strategy!” No really, we needed someone to tell us that.
You manage them by learning some basic financial concepts and then exploit them better than the competition. For example:
- Portfolio approach: a portfolio approach basically means that you “invest” in different kinds of opportunities so that if one doesn’t work out, the others can carry the weight. This is the classic way to manage the variability in influencer performance.
- Beta: beta is defined as the volatility (or variability) in performance. Beta sounds bad as it implies a lack of predictability but it also means that there is greater potential for upside. If you are in a “hits driven business” you need some beta. You’ll never make up for the losers putting out remakes of Jane Austen (assuming you can even get those to breakeven). You need moonshots. And yes, VCs also operate this way.
And here’s the thing: you might remember in the post on Pricing Based on Reach that we said that you should always estimate views on medians not averages precisely because of virality effects: they drive up averages in a way that distorts the expectations for a single video which is better represented by medians.
But once the content is published, virality works to our benefit: video views are bounded (limited) on the downside at zero. There is no similar bound on the upside which is where virality comes in. In fact, the video with the most plays in gen.video’s history didn’t come from any of the many multi-million subscriber influencers we work with. Instead, we got 11 mil views (which outs the fact that we haven't had a chance to collaborate with Mr. Beast yet. We came close but too many senior people at the brand hadn't heard of him. #facepalm) on a video from an influencer with 500K followers whose average video got around 100K views. And it was an add-on video that we tacked on to a campaign that was underperforming. What did we do right? We got ourselves up to the plate, as we do all day every day, and then got really, really lucky.
Which means this little pun is too good to pass up:
You can’t get a hit if you don’t swing the bat.
That said, we’d be remiss if we didn’t acknowledge what might appear to be a contradiction to all three of my loyal readers: in the post What Size Influencer Should You Work With we extolled working with fewer, larger influencers vs. tons of smaller influencers. If part of the game is trying to hit a winner, wouldn’t we be better having lots of bets on the table? Yes, to a point. This is the core reason that in a given campaign we work with 3 - 8 influencers vs. just putting the whole bankroll on a single influencer. We’re not so much banking on virality as mitigating downside “all eggs in one basket” risk while also making a reasonable bet that even with 3 or more, one is likely to nail it and outperform. I also believe the odds of virality are much better with a larger influencer and with all of the other costs described in the What Size and Hidden Costs of Influencer Marketing posts, the math ultimately favors the ‘intelligent, focused portfolio’ more than the ‘spray and pray’ approach.
What do you think? Is the percent of content that goes 'viral' high enough to mathematically justify using more influencers in a given campaign?