Paid-for bot followers, used to inflate influencers’ ever-increasing asking price, have received considerable attention this year. But behind the follower number are further wasted dollars related to reachability and targeting. A robust process is needed to maximize effectiveness and reduce waste.
Effectiveness of spend can be expressed as the conversion of dollars into impressions that increase propensity to convert or purchase. With the advent of swipe-ups in stories we can add more traditional effectiveness measures such as conversion attribution. For this piece I’ll be focusing on how to analyze effective impressions.
To date influencer rewards have been negotiated based on two factors: an influencer’s ability to reach an audience via posts and a brand match that clients believe will drive authenticity, therefore effectiveness. The follower count is seen as a rough estimate of the influencer’s ability to reach potential customers. The quality of brand match is presumed to drive the relevance of the reached audience, therefore the overall effectiveness of the investment funneled into the campaign.
Advertising effectiveness of an influencer is a function of three factors: the ratio of bot accounts to accounts owned by real people, reachability and the effective audience targeting. Methods to assess for all these factors vary in accuracy—in some cases brands will have to rely on subjective measures, in others data can be acquired and analyzed. Marketers work in a world of information gaps and of course work to close them to the best of their ability.
A way to gauge the effectiveness of a follower base of an influencer is to discount everything but the real, reached and desired audience–and only then establish the value of that audience. I want to build a case for working towards a CPM that is based on only the impressions of the reached and desired audience and then assessing influencers for their ability to deliver on it. By this logic we can work with available data and making best-effort estimates based on clues or industry averages where information is lacking.
Estimating the ratio of bot to real followers is an imperfect science. Monitoring a graph of followers over time can provide important clues however. Spikes without known causes, followed by a tapering off as Instagram slowly prunes inactive bot-looking accounts, strongly suggests a behavior of inflating the follower base by means such as outright buying of followers or engaging in programs that yield spikes of un-organic followers. Various data platforms around Instagram have popped up in the last 3 years that provide this follower graph, sometimes paired with sampling of follower profiles with data analysis that can tip off an algorithm that detects behavior that is typical of a bot account but not a real one. Working with data vendors such as HypeAuditor or with a platform can provide the important clues and provide an estimate for the bot ratio.
After discounting for bot followers one needs to consider whether a post reaches an audience or not. A follower, even if they are not a bot, may never see a post because they are either inactive on that day or the algorithm didn’t rank the post highly enough for it to be seen. This is reflected in the difference between the follower count and number of impressions in a campaign. Via the Business account program, Instagram has, since early 2017, started sharing historical data for posts with influencers and corporate accounts, both sponsored and organic. The number of impressions fluctuates depending on the whims and mood of the Instagram feed algorithm, which may trigger boosted distribution based on post performance signals (such as engagement or shares). Furthermore impressions may rise if posts surface in the Explore tab, becoming viral through message sharing or simply due to a paid boost by the influencer.
An influencers ability to convert from followers to impressions should hover around a base close to a historic impressions statistic average. Getting to such a baseline number requires a sample size of at least a few posts. Asking an influencer to share screenshots of the last 5 posts, both sponsored and organic, may be a prudent action. At Takumi we aim for at least 30% conversion from followers to impressions. Micro-influencers are proven to have a higher rate of conversion to impressions, which is why we optimize campaigns around smaller influencers where possible. The 30% conversion is based on data gathered from running 1,500 campaigns over thousands of influencers. Accounts with bots are bound to yield poorer results as bots don’t generate impressions, instead they just sit there inflating the follower count.
After discounting bot and non-reached followers a client still needs to assess the desirability of the remaining audience. Assessing the influencer’s aesthetic, interest, ethics and editorial are activities a brand can perform if they believe it to be a proxy for filtering influencers who they don’t believe to be reaching the audience desired by the brand. It’s a method, but a crude one, and prone to over-indexing for highly specific ideas of a brand persona. A client may instead, or as a provision, assess geographic reach and demographic statistics, both of which are made available through the Business account program and can be shared by the influencer in the form of screenshots from the Instagram Business account tab. Bots and accounts not reached can easily skew this data—something we frequently observe at Takumi in data from accounts with inflated follower counts. Bots may originate from Brazil or East Asia for example, whereas a high quality audience segment might be closer to the influencer’s home base. Data platforms go a long way with techniques not unlike those employed to detect bot accounts; providing demographic and geographic ‘fingerprints’ that can drive better decisions. Paired with brand intuition and screenshots from influencers, the worst of the flock may be easily pruned, or reward rates may be adjusted downwards to reflect the data.
All-in-all: combining a toolchain of tactics, data providers, data shared directly via screenshots and a tried and trusted talent pool, one can seriously drive the efficacy of influencer spend, already doubling every year.
Why influencer marketing to begin with?
The end product for influencer marketing is effective communication with an audience, not so much the influencers themselves. Influencers are collaborating with brands because they earn an income in exchange for the ability to reach people and weave a narrative to a degree of personalization unattainable to a brand that usually otherwise relies on mass communication for its marketing. It results in communication tolerated to a degree that traditional advertising today is not, perhaps even desired and on occasion appreciated. To an influencer it’s an ad support model. For an audience it results in highly relevant and relatable curation of both organic and sponsored editorial content. That an individual opts to continue to follow the influencers demonstrates the value and merit of this growing business model. This tension between the audience and influencer drives the self-correcting nature of influencer marketing: bad content or unsavory ads result in unfollows which diminish an influencer’s sway and ability to grow.
This symbiotic relationship that is the foundation of a nascent industry won’t work if clients continue to reward bot accounts, audiences that are not ever reached, and tolerate ineffective targeting. Of course advertisers have and continue to tolerate some level of waste in all channels because the effectiveness is presumed to justify the investment regardless. Influencer marketing isn’t any different. But the data and insight available are frequently disregarded or not factored into choices of marketers resulting in missed opportunities and unnecessary waste. The opportunity ahead for influencer marketing is to minimize said waste unleashing a virtuous cycle of larger commitments and more effectiveness. This gives more creative individuals an opportunity to subsidize their editorial and ultimately increases the level of personalization in advertising benefitting all consumers.
Leading managed service platform for influencer marketing campaigns, Takumi will be taking to the stage at Advertising Week – Europe on Monday 18th March at 11:45AM (Story Crafters Stage). Chief Revenue Officer, Adam ‘Sven’ Williams will provide guidance on how to activate influencers in interesting ways (beyond just a cookie cutter brief) to cut through the noise and win over consumers who are becoming increasingly frustrated by soulless product features.
Takumi is an industry leading influencer marketing platform that makes it effortless for brands to deliver successful Instagram campaigns across the full spectrum. Founded in 2015 by Solberg Audunsson, Mats Stigzelius and Gummi Eggertsson; the company has become the most active dedicated Instagram influencer platform, creating 35,000+ pieces of content, completing 1,900+ campaigns and working with 800+ brands. As experts in the field, Takumi provides a fully managed service which results in original branded images on Instagram and high-quality creative that can be repurposed across other marketing channels – bringing tech and in-house expertise together for a holistic campaign delivery.