Join below for the greatest social media newsletter ever. Ever Ever.👇
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
X
.svg)
It was 2016. Stranger Things had just premiered, and the Rio Olympics were in full swing.
Life was good for influencers and their brands.
But little did anyone know. Events were brewing that would upend the lives of influencers.
Cambridge Analytica was syphoning data from 87 million Facebook profiles using Facebook’s Open Graph and an unassuming third-party app. By April 2018, Facebook restricted access to Open Graph along with all its associated APIs – including the Instagram API.
That just so happened to be the API influencer agencies used to analyse and classify thousands upon thousands of photos. Brands could find perfect influencers instantly. Marketers and agencies could also pull accurate metrics and analytics. Not anymore.
Influencers, agencies, and brands were back to square one: dealing with muddy metrics and inconsistent results.
AI in influencer marketing, however, just might turn things around. (Hopefully without the scandal this time.)
Is influencer marketing an excellent opportunity to humanise your brand and reach new audiences? Yes. Is it riddled with problems that influencers and brands would bother rather not deal with? Also, yes.
Luckily, AI in influencer marketing looks promising for solving most of the industry’s problems.
Metrics, fake engagement, visual search, audience insights, even content selection – we might soon have a light at the end of the tunnel.
As of 2020, over two-thirds of brands report experiencing influencer fraud.
Fake influencers with artificially inflated engagement are everywhere. Unfortunately, there’s no consistent fool-proof system for analysing influencer engagement accuracy.
With Open Graph gone, brands have to use third-party AI tools. Newer AI-powered influencer platforms seem more accurate, but others aren’t. Some fail at separating genuine followers from fake purchased ones.
Some risks will always exist but fortunately, we’re moving towards a better place thanks to AI in influencer marketing analytics. That’s good news for authentic influencers and brands alike.
How can brands figure out if an influencer’s followers match their target demographic when they can’t even tell how many of those followers are human?
No one wants to address that one openly I bet. Fortunately, AI developers do.
AI and machine learning break the boundary that went up when Open Graph disappeared. Digital marketing tools can scan an influencer’s followers and spit out impressive analytics on demographics. Brands can make informed decisions on who to work with.
What if you could compare an influencer’s metrics and content to every other influencer instantly? That’s kind of how artificial neural networks work.
Tools with AI for influencer marketing take historical data from tons of influencers – either within your niche, network, or beyond. Using that data, the algorithm can decide what incentives are most effective for an influencer and when.
This might even include anonymised compensation eventually outside normal NDA clauses.
When AI platforms analyse an influencer’s presence, they look at the whole presence. The results are so impressive they’re hard to grasp.
Using natural language processing (NLP), AI can predict how well an influencer will meet a brand’s objectives for any campaign. No more trial and error.
AI can even use NLP to predict whether an influencer’s “influence” will improve or deteriorate and when.
Other industries already have access to copyright infringement algorithms and automatic notices. For influencer marketing, AI will level the playing field that Open Graph messed up here.
AI in influencer marketing will enable brands to automate the entire contract process in many cases. Later, AI can judge an influencer’s likelihood of breaking a contract or NDA before they even sign. And when both parties sign, AI can prevent infringements before they happen.
AI is bringing visual search back to influencer marketing in a massive way. Forget Open Graph. You already have machine learning tools for analysing photos across different platforms.
Whether you want images that contain your brand’s products, relevant themes, or competitor products, AI has you covered. It will even produce interactive reports based on other info so you have a comprehensive understanding of who fits best.
77% use Earned Media Value (EMV) to judge ROI from influencer campaigns. Do you know how to calculate EMV? Well, it uses historical data from prior influencer campaigns and it’s pretty complicated.
That’s why metrics and ROI are the best places to improve influencer marketing with AI.
AI and machine learning can analyse potential ROI from hundreds of dynamic angles: influencer-brand combo, seasonal trends, benchmarks, content, and more.
Plus, these metrics and ROI estimates are persona-based so they automatically adapt for influencer marketing’s uniqueness.
Sick of A/B testing what kind of content performs best across each influencer’s audience and meets your brand’s objectives? AI can help your brand and influencer decide the best content solution at scale.
You didn’t think AI in influencer marketing was all about metrics and data, did you? In fact, virtual influencers are (not-so) alive and well.
We designed one for essence Cosmetics in 2019 called Kenna. At the time, KFC, Samsung, and other brands had already launched their own virtual influencers too.
Does this mean the world is ready for human influencers to hang up their hats? Not exactly. That certainly wasn’t our mission with Kenna.
What does it even mean to “work with” virtual influencers anyway?
Some virtual influencer creators would say it’s a unique opportunity for brands to create a deep emotional connection with their audience. That sounds a little dystopian but there are some pros and cons to consider.
Pros:
Cons:
What are the implications of using AI-powered analytics to propel engagement for virtual influencers? It’d reach a tipping point eventually where virtual influencers push humans out of our feeds. Sounds like the closest we’ve come to singularity, no?
It’s almost frightening how quickly virtual influencers manage to grow their accounts. Some have over 20 million followers across all platforms.
But they’re not all creepy. Cartoons and characters make AI and virtual influencers fun and light-hearted. Here are the top 5 of 2021 based on engagement and followers:
You don’t even have to worry about virtual influencers replacing humans because you have so many other opportunities to use AI in influencer marketing.
Specifically, these influencer marketing platforms can help close the gap (and then some) after losing Open Graph access because they’re powered with AI.
Linqia uses AI to power every piece of its influencer marketing platform. Brands can search for influencers with its AI content analysis and optimise campaigns.
Maybe most importantly, Linqia uses closed-loop measurement to track engagement analytics by linking with data from other relevant sources like smart TVs and in-store shopping.
This translates into more accurate and comprehensive analytics. No wonder McDonald’s and others already used Linqia for campaigns.
Influential is a powerful AI platform with natural language processing from IBM Watson. The Influential website is super minimalist with nothing but some basic service terms like “media strategy.”
Farther down the page Influential displays a series of brand logos and case studies for huge companies like Pepsi, Venmo, and Wells Fargo. Even the case studies mention little more than metrics so it would seem Influential has something huge on their hands here (for those willing to pay).
TikTok offers an impressive built-in Creator Marketplace. Brands can search the platform’s influencers by keyword or category and filter according to their specs to find the perfect creator for the job.
All you need is a TikTok business account to search. Best of all, the Creator Marketplace runs on TikTok’s first-party metrics. Brands can feel confident in an influencer’s engagement numbers and that the influencer’s audience reflects what the brand needs.
When we talk about AI in influencer marketing (or anywhere), it’s really the machine learning that makes it so impressive. That’s what allows the AI system to accept new information and learn to improve its output.
For influencer marketing opportunities, that new information could include millions of photos, likes, comments, followers, videos, and more. Using this tech to reduce fake influencers, inflated engagement, and inaccurate metrics will create a better landscape for both influencers and brands.
Got an idea for using AI or virtual influencers in your next campaign? Check out our work on Kenna for essence Cosmetics for inspiration.
As an Associate Strategy Director I am responsible for leading the development and execution of organic and paid social media and content strategies for our clients at Kubbco.
These might interest you...
why should I sign up for the newsletter?
