Data-driven targeting could decide the brand of your next toaster – or the name of your next prime minister.
6 minute read
Let’s go back a few years to 2016 and a company called Cambridge Analytica who used the personal Facebook data of 50 million users to help the Republican Party (led by Donald Trump) embark on a highly sophisticated and targeted digital marketing campaign.2
Instead of expensive national TV campaigns, budgets were re-directed to social media where, in hindsight, ROI was impressively higher.
The idea was to map personality traits based on what people had liked on Facebook, and then use that information to target audiences with digital ads. To do this, Cambridge Analytica used details on users’ identities and ‘likes’. But what was perhaps most effective was the use of look-a-like audience targeting, which opened up potential new bases in the hundreds of thousands in a few deft clicks.3 The value of this tool was not lost on Trump’s senior adviser for data and digital operations, Brad Parscale, who stated: ‘One of the most difficult tasks of a political campaign – distinguishing likely supporters from the undifferentiated mass… can now be accomplished instantly through artificial intelligence.’3
It also cannot be ignored that group-based targeting is simply more effective at achieving certain objectives. Persona-based targeting often ‘ignores the fact that people are social creatures [who] belong to communities and are part of influence networks that they use to decide what to watch, read, buy, and pay attention to’.4 Where targeting based on ‘likes’ may be all that’s needed to nudge a new toaster into your cart, who you choose to politically align yourself with is often informed, or reinforced, by your broader social or cultural context.