Brands that have issues with influencer marketing may have themselves to blame
Influencer marketing has risen in popularity over the years, driven in part by driving sales, quality content creation and third-party credibility. But as the industry has grown, some issues have emerged. Three of the major ones have been caused at least in part by how brands themselves go about working with influencers. All three can be easily avoided:
1. Fake Followers/Fake Likes
One of the significant values of influencer marketing is the audience that the influencer brings, so naturally, a larger following was worth more to brands. It didn't take long for at least some influencers to realize that buying cheap followers (as low as $50 for 10,000) could make them more money.
The problem that brands later realized is that engagement rates fall as (real or fake) follower counts grow. Microinfluencers, which we define as those with roughly 20,000 to 200,000 followers, typically have the strongest engagement rates.
To bypass fake followers, some brands started paying for "performance" by compensating based on likes and comments on influencer posts, assuming this engagement was a sign of content quality. That quickly led to formalized buying, selling and trading of likes and comments.
We've found, over the course of our programs, that either flat-fee compensation or sales-based compensation (such as affiliate models) best remove the perverse incentives brands have created that encourage influencers to artificially inflate their numbers.
2. Saturation Rate
This metric hasn't gotten nearly the amount of attention as fake followers, but it's arguably more important. Last September, I defined saturation rate as the percentage of an influencer's total content that is sponsored.
What we've seen is that as the saturation rate climbs, the influencer's engagement rate plunges. Why? People follow influencers for inspiration, discovery and keeping up with trends. If all of their content is paid for by someone else, their influence disappears more quickly than their reach.
How did brands help create this problem? By not routinely checking this number. While a good influencer creates content on time, does beautiful work and is generally a pleasure to work with, that can mean that brands flock to her again and again, crowding out any time she has for her "regular" content.
I don't begrudge the influencer for getting paid, but brands need to remember they are looking for influence, not just reach and not just attractive content.
3. Non-Disclosure Of Sponsorships
The FTC has been very clear for years now that "meaningful connections" between brands and influencers must have "clear and conspicuous" disclosures. That rule, however, is openly flaunted by brands and influencers alike.
A recent article in the Harvard Business Review noted that 28% of influencers were asked by brands to not disclose their connection to the brand. Why? Out of the belief that disclosures will damage the credibility of the content.
What's interesting about that rationale is that the same article notes that consumers are now equally likely to make a purchase regardless of disclosures. Some may even believe that the official #ad disclosure means the selected influencer knows enough about the industry to be hired by a brand and is, therefore, credible.
We've been able to track significant sales lift coming from programs with compliant disclosures, so in our experience, the fear of loss of credibility is overblown. Really good content thrives despite disclosures. Average or below-average content may be further degraded by disclosures, but a good influencer marketing program separates the high-performing content from the rest anyway.
As brands set their 2020 influencer marketing plans, the lessons learned from the past are easy to incorporate. By paying either flat fees for assignments or employing affiliate compensation models, by checking influencers' saturation rates and moving away from the worst offenders, and by demanding clear and conspicuous disclosures, the industry can continue to improve.