The Impact Of Artificial Intelligence On Influencer Marketing

In October 2017, Facebook altered the Instagram API to make it harder for users to search its giant database of photos. The change was a small element of the company’s response to the Cambridge Analytica scandal, but it was a significant problem for parts of the digital marketing industry.

Not long before, New York-based influencer marketing agency Amra & Elma had developed a platform that ingested data from Instagram, and allowed its client to use AI image classifiers to find very specific influencers. For instance, they could find an influencer with, say, between 10,000 and 50,000 followers who had posted photos of themselves in a Jeep. Facebook’s move killed this capability in a keystroke. Another day in the digital duel between the AIs deployed by digital marketers, and those deployed by the social media platforms.

Once upon a time, successful advertising was a question of artistic creativity and financial muscle. Hire a hot agency to produce a 30-second ad, spend a lot of money with TV and newspaper owners, and you were bound to make a splash. The Mad Men of Madison Avenue were about sizzle and silver, not science.

The internet has changed all that. Today it is all about data, analytics, and algorithms. Google and Facebook have made off with much of Rupert Murdoch’s ad revenue, by the simple method of making adspend accountable, and much, much more targeted. With Google Ads you only pay when a consumer clicks on your ad. With Facebook ads you know far more about the type of people your ads are being served to. It is no longer “spray and pray”.

Digital marketing is now the sexy part of the advertising industry, and its currency is data. The ability to obtain and analyse data is the core skill, not a good line in patter and a generous expense account. It is more important to understand the mechanics of the social media platforms than to know the maitre d’ at the latest trendy restaurant.

Elma Beganovich, co-founder of Amra & Elma, says “AI is a game changer in digital marketing, especially influencer marketing. Brands are now able to pinpoint their target demographic precisely, and speak to them directly. Technologies like voice recognition allow us to identify what topics influencers are talking about on YouTube and Instagram TV.”

Influencer marketing is a significant and fast-growing section of the digital marketing industry. The big agency networks (WPP, Omnicom, Publicis, and Interpublic) have their players, and there are hundreds of independent outfits. The industry is currently valued at $8bn, and before COVID-19 came along, was forecast to double in the next two years. Jessy Grossman, founder of Women In Influencer Marketing (WIIM), thinks that the virus might accelerate growth: “Brands have been hit hard. We are going to see them pinch every penny and demand the most ROI. Digital marketing is far more efficient, data driven and quick. It’s also the most economical way to get production quality content and large-scale distribution at a fraction of the price.”

In its early days, when Instagram was new, influencer marketing was a lawless, wild west environment – the least data-driven part of the digital marketing industry. Clients were eager to benefit from the adulation that influencers can inspire, but they had no way to measure the return on their investment, and unscrupulous influencers were buying fake followers by the thousand.

But analytics are taking over, even here. Clients want to know what they are getting for their money, and consumers want to know when they are receiving sponsored messages. Analytics are critical in understanding the level of engagement between influencers and their followers. And regulators have corralled influencers into using hashtags to indicate when a post has been paid for.

Influencer marketing is still mainly a B2C (business to consumer) activity, with the big three client categories being beauty, food, and fashion.