Is the golden era of online political micro-targeting ending? The recent revelations about how data analytics firm Cambridge Analytica exploited Facebook data, extensively documented in The Guardian, Channel 4 News, and other major media, seem like a watershed moment for this industry.
Much of the story is driven by Canadian-born whistleblower Christopher Wylie, who was a founding member of Cambridge Analytica and has now come forward to expose their inner workings. Their firm specializes in collecting and processing large amounts of consumer data for the use of political marketing online and Wylie claims that the data set first used by him and his team to create the company’s touted “psychographic” targeting profiles was collected without user consent – up to 50 million Facebook users.
The Globe and Mail released a great summary of the scandal, so we won’t repeat the technical and legal details here. What we’re more interested in addressing is whether any of this is new in the realm of political marketing.
“Micro-targeting” and “big data” have been buzzwords in political circles for the past decade and no one should be surprised that nefarious actors would try to exploit the holy grail of accessible big data that is Facebook. While Cambridge Analytica appears to have collected its data through unethical means, the techniques and processes that resulted from analyzing that data are not inherently new. Pollsters and communications specialists collect data and match key messages with target demographics on a daily basis; Cambridge Analytica simply created a more sophisticated and streamlined version of that process by using its algorithms and having access to a potentially priceless dataset. So whereas you and I can deduce what demographics would be susceptible to pro-gun and anti-gun messaging, Cambridge Analytica appears to just do it more granularly.
What intrigues us is how they were able to use that targeting and execute so effectively. The operations and workflow planning, if you will. In our experience, it’s rare for a lack of data to be the main roadblock for sophisticated political advertising online – it more typically comes down to a lack of available resources and/or coordination. Wylie says in his interview “In addition to having data scientists, psychologists and strategists, they also have an entire team of creatives, designers, videographers, photographers. They then create that content, that then gets sent to a targeting team, which then injects it into the internet.” Here’s a direct link to the quote. The point is this: Cambridge Analytica’s suspected massive influence doesn’t just come from having access to better data than others, it’s largely (and in our view, more so) to do with the fact that they appear to have learned how to efficiently and ruthlessly produce inflammatory content to “weaponize” that data.
In addition, its clear from Channel 4’s undercover interview with CEO Alexander Nix that his company has no qualms about distributing content on Facebook under falsely identified Pages – something that “the Russians” have been extensively accused of doing – and what should run contrary to most election laws. Did they do this in the United States, and break electoral financing and disclosure laws? We don’t know – but we can tell you from our experience working on the Facebook platform that the technical barriers to doing so were, up until recently, essentially zero (Facebook will tell you this is about protecting freedom of speech and keeping the internet clear of censorship, and they are right… to a point).
So, where do we all go from here? First, it seems to us that regulators are now finally starting to catch up and understand how data is collected and distributed on Facebook, and what this means for privacy. This will likely lead to updated laws and enforcement mechanisms. Second, elected officials are rapidly cluing in on how serious a threat to democracy the exploitation of Facebook’s data poses, and will certainly start placing more checks and balances on online advertising. Third, Facebook is already taking steps to fight false news and improve transparency, such as with its new advertising transparency tool, and will continue to change – probably very fast. Fourth, we think that Facebook’s days of absolute dominance of Western social discourse, are likely (though not certainly) numbered, and we are increasingly expecting a new serious entrant in the market or a resurgent of a secondary player.
So yes, we do expect that to some degree, the era of cheap and easy online political targeting is coming to an end. And we think this is a good thing. But also, in our view, we as public affairs professionals should not forget where we came from – before the era of micro-targeting and Facebook – and starting planning to return to our roots.
By this, we mean that good messaging – authentic, creative, compelling storytelling – should and in our view will always be more important for politicians than massive data sets and overly micro-targeted campaigns.
So, we say to Cambridge Analytica: (hopefully) good bye, and (certainly) good riddance. There is a better way to do politics online than a race to the bottom of the dataset. Let’s work towards that.