Will you lose your Influencer marketing job to A.I ?
January 30th 2018
I'm pretty obsessed with A.I, i'm studying machine learning coding and have my head buried in A.I books and journals. Recently, i've been reflecting on how it will affect marketing and strategy in general and myself as an Influencer and culture marketer specialist.
Motivating individuals to endorse a product is an age old goal of marketing professionals. While previously the goal was to ‘hook’ a celebrity, influencer marketing brings with it a whole raft of new challenges. There are far more potential Influencers directly interacting with those that could be influenced. And marketers have relatively little control over messaging, and even less control over the messages that end up ‘viral’ - for better or for worse.
There are lots of tools that search by interest and demographics, but Emerging AI neural networks may be able to help cut through the noise and pick out the right kind of Influencer and even manage the relationship - at least in some capacity. They can effectively identify Influencers and micro-influencers, shed more light on how to improve messaging, and even improve the odds of creating a viral campaign.
In essence, neural networks are AI systems that are ‘trained’ over time to recognise patterns and make connections in order to complete simple tasks, or more usefully in marketing, analyse data.
With over 2.5 billion worldwide social media users and many Influencers having hundreds of thousands of followers, there is so much data available on the effectiveness of Influencers and messaging that it is almost impossible for a human to comprehend. But it is relatively trivial for an AI system.
Identifying Influencers is currently a somewhat crude process, involving follower numbers, raw volume, and perceived fit-ability, there are some tools that claim to help - and even claim they use A.I. But a proper A.I neural network, one that shows the patterns and behaviours of Influencers who have been proven to be effective, (and with the right access to an API) can easily spot similar users without relying upon the underlying metrics. This opens up an entire segment of smaller, but highly engaged, potential Influencers.
Neural networks will also be able to instantly spot, and therefore avoid, ‘fake’ Influencers - the accounts with a large percentage of bought followers or ‘bot’ shares. These accounts tick all the boxes metric-wise, making them easy for humans to believe. But the usage patterns are so far removed from those of the effective real Influencers that they will not fool an A.I system.
Identifying messaging effectiveness in the age of social media is also difficult for humans. We can see direct shares and even catch some degree of wider spread, but the majority will be lost. This makes it difficult to see how effective, or ineffective, a strategy actually is. Neural networks can sift through data across platforms looking for the same, or very similar, messaging, giving a great deal of information about the true effectiveness of a message and the true reach of an influencer.
In the end, I think A.I automation will help brands save time and be more effective, presenting relevant Influencers who have a genuine connection to a brands purpose. It should also be an incredible tool that technically literate marketers can use to dramatically improve the reach of their messaging. Hopefully, this will give more time to focus on creating well thought out strategy and integrating with culture, and it may even increase the likelihood of creating a viral campaign - bring it on...