Vivien Garnès is an entrepreneur, currently co-founder and CEO of Influencer Marketing SaaS company Upfluence Inc, one of the world leaders in the industry, headquartered in New York City, venture backed, with almost 100 employees.
He defended his Executive Doctorate in Business Administration (Executive DBA) thesis in September 2021, on the topic of ‘Return on Influencer Marketing Investment’, under the supervision of Professor Kiane Goudarzi, full Professor at iaelyon School of Business, Director of the MBA programme, Deputy Director of the Magellan Research Centre and DBA Site Manager for Business Science Institute in Shanghai.
Thesis Direction
Pr Kiane Goudarzi
Thesis Title
Return on Influencer Marketing Investment
Abstract
Influencer marketing is a rising star in the field of digital marketing – (i) credited to be a $20B industry by 2024 and (ii) growing rapidly at 32% CAGR (Markets & Markets, 2019) – (iii) while the term itself seems to have been coined circa 2004, just 17 years prior to these lines being written.
Still, looking at 7,400+ online sales generated by social media influencers, representing $5.7M+ in gross transaction value (GTV), there is no statistical dependance between influencer’s costs and the Return on Investment they yield (a correlation of -0.0294). The sad truth is that marketing managers do not spend their budgets profitably.
Yet, Academia has, unfortunately, not done much better. While the topic of influencer marketing has been significantly written about – especially “why” influencer marketing impacts the consumer decision process psychologically – nothing reliably discusses “how much” it impacts influencer marketing ROI (ROIMI). We are left with (i) proposed alternative metrics (EMV, AVE, SMIV), which do not quantify costs or sales and are the ultimate admission of guilt, and (ii) grey literature based on declarative survey data at the budget or campaign level, which is not insightful nor actionable for practitioners.
Emboldened by the gap in the literature, and by the room for improvement that’s emerging from the data, the goal of this research is to offer marketing professionals an understanding of (i) the terminology behind ROIMI, (ii) a proposed method of calculation, (iii) which variables improve ROIMI and (iv) how to predict influencer costs to sustain high ROIMI.
Having access to a large dataset of observed influencer-generated sales (tracked via coupon code usage), the researcher carried out a quantitative analysis based on multiple linear regressions—delivering a structural equation model, mapping variables that impact ROIMI.
While ROI directly is hard to predict, the model reveals that two variables should be prioritized to maximize orders: (i) influencer follower count, (ii) influencer comment count, whereas (iii) influencer like count negatively impacts it. Interestingly, influencer follower count, and influencer comment count also increase influencer costs, as it is the legacy way influencers and brands have determined costs, hurting ROI as a result.
In order to challenge this status quo, the model was then broken down into the following influencer cost predictive inequation, to give practitioners the means to sustain high-ROIMI campaigns:
inf_cost < (4.575e-05 * inf_follower + 0.028 * inf_average_comments – 7.798e-04 * inf_average_likes) * Average_Order_Value / Expected_ROI
Running this influencer cost prediction against the initial dataset generated a 50.28% improvement on ROIMI while maintaining a constant budget. Certainly, a great testament to the model’s relevance.