Reflecting on my previous blog post about research into influencers and their impact on purchase intentions, and the generally supportive reaction to the main argument of the post, that Twitter’s so-called research was not only self-serving and entirely without real content, but was also based on flimsy, at best, research practice, I started thinking about how influencer marketing has become something of a totem for digital marketing companies looking for the next big thing to foist on susceptible and perhaps untutored clients.
The research conducted by Twitter and analytics firm Annalect (for what it’s worth: no market research skills or expertise is shown on their website), bigged up the supposed role of “influencers” in driving consumer purchase intent. The aim of the research was to identify the percentage of Twitter users saying they’ve made a purchase as a direct result of a Tweet from an influencer” (http://bit.ly/1TcorFK). The only thing surprising in the 40% of respondents reported as confirming this is that the number isn’t actually higher. Seems an oxymoron that an influencer isn’t much of an influencer if they don’t influence!
The research, a prime example of headline-generating PR tactics, was presumably aimed at promoting Twitter’s Niche service, which, just coincidentally is an influencer marketing programme and platform… The strategy’s clearly worked, at least insofar as it generated a few headlines, as exemplified by AdWeek’s: “Twitter says users now trust influencers as much as friends”, and some uncritically admiring tweets.
Whilst my initial posting on this topic focused on the lack of transparency in the research method and the paucity of any new (or even old) insight in the findings, it got me to thinking about a more substantive issue – one that has been nagging away at me for some time: how to create an accurate, rigorous and actionable metrics framework to identify, track and measure the impact of so-called influencers in the social media sphere. There’s no doubt that this is an important priority for communicators – identifying influencers who genuinely influence would certainly allow for the development of a new effective channel to reach key target segments. I should stress that this process of identification needs to be treated distinctly from how influencers are themselves influenced – for example, it’s good to see the increasing debate in the PR media about the ethics of paid influence.
The role, importance (and identification) of influencers in triggering and driving consumer behaviours, whether purchasing consumer goods, making healthcare choices, voting decisions or donation giving, has been a matter of debate for many decades. I was first made aware of the importance of influence via John Gilfeather, a distinguished researcher and expert in all things reputation, who insisted I read his colleagues, Ed Keller and Jon Berry’s book on Influentials (published in 2003, but still in print, and a recommended read). The book is a serious, systematic attempt to understand the power and reach of Word of Mouth. Their findings are based on firm research, from which they derive their key finding that there is a relatively stable ±10% group of the US population that directly influences consumer and peer group decision making and behaviour.
Pre-Twitter, of course, it was easier both to define and also identify “influentials”, as Keller describes them. Influencers were identified predominantly as those with a community voice – politicians, educators, community leaders. Peer group influencers tended to be relatively easily identified, with a stable, and long standing impact on their group. Influence also tended, back then, to be communicated either by personal interaction (word of mouth) or via editorially mediated channels – TV, radio, newspapers and magazines.
How things have changed, driven by the explosion in social media platforms and use. I recall a senior Hewlett Packard exec arguing at a conference some years ago that Facebook was becoming the dominant search engine driving purchase preference (over Google). Whilst that executive may have got the specifics wrong, it was nonetheless a powerful insight that social media would begin to play an important role in purchasing behaviours. Social media platforms generally have become powerful intermediaries in consumer decision making – at least as important as “traditional” sources of influence as detailed in Keller’s research.
In a world where millennials and post-millenials increasingly consume their media from a multitude of sources, mainly funnelled through their phones or tablets, the more or less stable identify that characterised the influential is no longer a given. A constant swathe of new, filterable, micro communication channels, coupled to the decreasing attention span of these youthful audiences means that influence, and influencers wax and wane. However, one constant is the role of “friends” as a key driver of word of mouth, which is arguably still the strongest form of behaviour-driving influence. Nothing surprising in this, of course – our friends are friends precisely because of the emotional connection we share with them. And there is increasing evidence that our behaviours and decisions are predominantly driven by our emotional states (i.e. our stomachs), rather than rational cognition (our heads). * So clearly, friends, by virtue of their emotional connectedness, will undoubtedly have a direct impact on decision making.
The challenge for those of us with an interest in raising the standards of measurement, is to lift the debate above tautological, vapid headlines that simply assert the importance of influence. Measurement and analytics specialists need to work alongside clients who are willing to reverse the measurement of influence, away from meaningless metrics based on followers, likes, shares etc (and calling these metrics or variants “engagement” doesn’t make them a measure of anything other than brain to mouse movement), and to recognise that the challenge of measuring the impact of influence and identifying true influencers requires rigorous research into the “last mile” – i.e. what drives behaviours in the real world, rather than what we extrapolate from within the Twitter, Facebook or other social media ecosphere.
I don’t think there’s anything spectacularly controversial in any of this – the Barcelona Principles, for instance, have always stressed outcomes rather than outputs as the key determinants of any truly valuable measurement framework and programme.
However, there are signs of a more robust approach being discussed. It’s encouraging that a number of sceptical voices have been raised arguing that influencer marketing is not the golden bullet the PR industry has been looking for. In truth, measurement success is, and always has been, really all about the data, properly collected and then analysed in a rigorous and consistent manner, and without promising to deliver more insight than its metrics can support…
* The role of emotions in driving behaviour is a fascinating, and important topic, with potentially enormous significance for marketers. TMP will be publishing a White Paper shortly, detailing the history, development and current state of emotional research, and its application within communications and market research.