Influencer matchmaking AI is not the silver bullet it claims to be

In continuation of my opinion piece titled ‘How can influencers ensure they are not held hostage by platforms?, it is pertinent to mention the new technologies, especially artificial intelligence (AI), that are at the disposal of marketers and branding firms to enhance the promotions and brand equity of companies. 

Artificial intelligence is demonstrated by machines when they display analytical skills in more efficient ways than they would work through human intelligence. AI influencer marketing systems use Artificial Neural Networks (ANN) to analyze millions of online images and videos within a few seconds. This, in turn, helps identify images with certain attributes or characteristics such as a particular brand logo or product.

In this developing technological time where people tend to have millions of followers, likes, and views on their Instagram, Facebook, and YouTube pages – and with similar platforms like TikTok, Bigo, and SnackVideo being downloaded in millions per day – it is commonly known that the best way to reap profits for any particular brand, product or service is to precisely and accurately target the audience through influencers. However, instead of having to do this research by feel, AI platforms simplify this process and help brands match with influencers having similar traits in a very less time. 

AI systems can collect data for marketing firms and then have them narrow down the influencers they want to target or have the brands be endorsed by them. This helps them in identifying unique and upcoming trends, trace the existing ones, and manage to monitor their growth strategy according to a very specific plan. Influencers can be evaluated on the basis of how much reach they have, and brands can thus calculate how much potential investment can be further made.

INCA by GroupM is among the various Pakistan based solutions such as Bradri, Den, Walee, and Amplifyd are now in either the development phase or have already started working through this AI system. This AI system helps in identifying the fake followers and engagement done by the influencers as well which also gives the marketers that edge over traditional methods. This can be identified by inflation of followers, fake engagements, fake posts. AI helps in identifying these loopholes.

With great power comes great responsibility—one of the popular phrases of Stan Lee comics Spiderman, the firms have gotten their hands on a genie of Aladdin or a wishing dragon which they believe is solving their problems. Be it gathering the data, cross-referencing it with other social media platforms and followers, filling up numerous spreadsheets and dashboards and recommending in seconds on the strategy to be used to target and push its brands.  

With so much work done in so little time and with so much ease, marketers and firms are now relying heavily on this AI system. In fact, it can be considered as a cure to a lot of problems in the digital world. However, Influencer matching through AI is not the silver bullet it claims to be. 

It is just a solution that is currently working and is a one size fits all approach thus getting desired results sooner than later. Why is this so? Firstly, because the influencers who have smaller follower numbers mean that the influencer has more of an effect on each individual follower than having millions of inflated numbers to meet the target. There are bot generators that produce and generate inflated and fake numbers. 

A ‘bot’ is a software program that performs automated, repetitive, predefined tasks. Bots typically imitate or replace human user behavior. Because they are automated, they operate much faster than human user. Secondly, and most importantly, an AI system is a learning system and it keeps on updating itself and the data that is being fed to it. The more refined the data and the more accurate procedure the marketers follow and the better results they will get.

What really matters is that due to AI, the elements of human interaction, and emotion – the traits that help marketers identify what influencer would have the best association with their brand or product – are taken out of the equation. The audience sentiment, the overall environment and the creation of the brand is hampered. 

Further it requires dedicated resources and human capital to study the influencers, their routine, habits, the product or service to be endorsed, the target market and then calculate the return on investment (ROI) will take time in this quick result oriented field.

Then comes the brand and quality of product service issues. If the influencer matches the brand as per the AI system, the number shoots up but later on due to lack of quality of that product the same influencer loses the number the AI system would automatically learn and adapt on the data it gets. 

Irrespective of the quality of the product and the endorsement of influencer which was not under his/her domain but the loss of numbers is easily picked up by the AI system thus showing skewed results. The loyal followers or genuine numbers would always tend to stick with the influencer and be responsive on their posts whether it’s Instagram picture, YouTube or TikTok video. 

In no manner does it means that benefits of technology cannot be utilized, only a machine can deal with the mass amounts of data and inputs and give top-level insights, the whole question revolves around how well the marketers and the firms improvise on the directions and instructions that they give to the AI system as all programs are only as good as the instructions it is provided and that is through a human to help it and guide it.