Friday, 3 June 2016

Are influencers overpaid? A data-driven exploration of influencer metrics

A recent article in Digiday made the bold claim that influencers are overpaid and generally do not advance business objectives. Christopher S. Penn examines the merits of this claim for SHIFT Communications. Follow SHIFT Communications on Twitter: @SHIFTcomm 

WHAT ARE INFLUENCERS PAID TO DELIVER?

The first, most important question to answer is what are we buying when we work with influencers? Some influencers and their agencies make vague, illusory claims about what their influencers bring to the table. What do we receive when we pay influencers?
Influencers are paid to deliver:
  • Awareness: influencers are paid to deliver net new audiences to brands, audiences who might not otherwise be aware of the brand.
  • Trust: unlike first party advertising, influencers are paid to bestow additional credibility on a brand’s products and services.
  • Engagement: influencers are paid to create audience engagement with a brand, especially on social and digital channels.
  • ROI: influencers are ultimately paid for business impact, and their business impact must be measured against other advertising channels.
Influencer marketing is not new; celebrities have been doing endorsements for decades, ever since the invention of mass media. Who those influencers are and what channels they operate on changes constantly, but the fundamental principles, the fundamental deliverables have not changed.

DO INFLUENCERS DELIVER?

Do influencers deliver? How would we know if they were delivering what we pay them to deliver? The answer to this question depends on analytics. Any influencer who refuses to provide complete first-party data (such as their social media analytics) should be avoided, as they have something to hide.
Even without their cooperation, we marketers and communicators can see a tremendous amount of information. What can we access?
  • Awareness data: we can see their total audience reach, their total following across channels. While not the end metric, audience data is a valid starting point.
  • Engagement data: we can see how engaged an influencer’s audience is. We can identify social engagement activity such as likes, comments, and shares.
  • Clickthrough data: by mandating the use of an open click-tracking service such as bit.ly (disclosure: a SHIFT client), we can identify how many audience members click through to digital properties we care about.
  • Brand baseline data: we know our own analytics. We know what reach, engagement, and action rates we can generate; if we compare our baseline data to a potential influencer’s data and find that we generate audience, engagement, and action rates orders of magnitude greater than a potential influencer, we know they’re not worth paying.

A BRIEF SURVEY OF THE INFLUENCER LANDSCAPE

Using our marketing technology systems, SHIFT extracted influencer data from over 16,000 different influential individuals on Facebook, Twitter, and Instagram, totaling 3,640,894 social posts across all platforms. The timeframe for this examination is April 1, 2015 – March 1, 2016. You can download your own copy of the Tableau workbook off Tableau Public.

MEDIAN AUDIENCE SIZE

We examined the median audience size on each network of our 16,000 influencers:
  • Instagram’s median audience size was 258,301
  • Facebook’s median audience size was 7,797,648
  • Twitter’s median audience size was 108,064
Median Audience Size

ENGAGEMENTS

We examined overall engagement numbers per network for 16,000 influencers:
  • Instagram’s median engagement of likes and comments was 3,705 per day
  • Facebook’s median engagement of likes, comments, and shares was 5,101 per day
  • Twitter’s median engagement of likes and retweets was 10 per day
Total Engagement
In terms of engagement rates, our influencer pool generated:
  • Instagram median engagement rate of 1.44% per day
  • Facebook median engagement rate of 0.10% per day
  • Twitter median engagement rate of 0.01% per day
Engagement Rate
In terms of shares/reshares – a metric which tells us how much extra marketing an influencer is doing for us, our influencer pool generated:
  • Instagram median share rate of 0; Instagram doesn’t have resharing capabilities built into the platform
  • Facebook median share rate of 0.019% per day
  • Twitter median share rate of 0.031% per day
Share Rate

HOW DOES INFLUENCER PAY COMPARE?

The best benchmark to understand what’s fair for influencers to be paid is to look at other ways of generating the same metrics.
  • If we care about audience growth, we pay a social network to generate those followers for us from a targeted audience.
  • If we care about engagement, we pay for engagements from a social network’s targeted audience.
  • If we care about content acceleration, we pay a search or display ad network to show advertisements to our audience on relevant properties – including influencers’ blogs, if they accept advertising.
  • If we care about sales, lead generation, or ROI, we measure the cost per acquisition or increase in purchase intent from a well-managed, well-tracked influencer campaign versus other marketing channels.
Marketing strategy is about choices; if an influencer costs less to generate a specific metric than another advertising channel, use influencer marketing. If other marketing channels and tactics deliver better results, we choose those instead. Whether we overpay or underpay is a red herring question; the real question is, are we purchasing the desired outcome in the most cost-effective manner?

BEST PRACTICES FOR EFFECTIVE INFLUENCER MARKETING

Given our data and analysis, what can we recommend for brands interested in influencer marketing?

INSIST ON MEASUREMENT

The most important single thing we must ask for are metrics and analytics data. Beware any influencer who refuses to provide data; chances are they have something to hide. Give added preference and consideration to influencers who understand data and are willing to integrate their data with ours.
Rare but special would be an influencer who installs our tracking tags on their digital properties for the duration of a campaign; should we find such an influencer, we should consider compensating them at a premium rate.

MEASURE AS PART OF INFLUENCER SELECTION

Use defined criteria for influencer selection. As we can see above, we have access to audience size, engagement, and other metrics of influencer analysis. Use this data in advance to help choose influencers, rather than wait for potentially lackluster campaign results.
When choosing influencers, decide on the metrics that are most important. If net new audiences are most important because our product marketing has saturated our existing audience, we should choose influencers with an emphasis on reach.
If loyalty and purchase consideration are most important, we should choose influencers with an emphasis on engagement.
If sales, lead generation, and cost per acquisition are most important, we should choose influencers with high rates of action taken on their content; influencers who wish to be considered for such campaigns must provide analytics data and access.

MAKE INFLUENCERS WORK FOR YOU

One phenomenon we’ve experienced at SHIFT in our work with influencers is that influencers very often use brand dollars to promote their own channels and content. This has the net effect of making the influencer more powerful, but has not typically resulted in significant benefits to the paying brand.
Ensure that account/handle takeovers are a mandatory part of influencer programs and insist that influencers promote the brand handle with specific details about their takeover to ensure maximum impact.

CONSIDER PAY FOR PERFORMANCE MODELS

Consider holding influencers accountable based on strict performance metrics. Offer scaled compensation based on established metrics such as reach, engagement, clickthrough rate, or even ROI, using affiliate marketing software.
Influencers who generate real, reliable results should be happy to earn extra money above and beyond a base pay; influencers who have something to hide will generally avoid being held to any kind of metrics.

CONCLUSION

At the end of the day, whether influencers are overpaid or not is something of a moot point. What matters most is whether we know and measure what we’re buying from influencers. If we treat them as we treat every other marketing and advertising channel, we can make rational, data-driven choices about why, when, and how to use them to deliver maximum marketing communications results.
If you’d like a data-driven exploration of the influencers you’re working with or considering working with, contact SHIFT today. We’re happy to help.
Christopher S. Penn
Vice President, Marketing Technology


No comments:

Post a Comment