Tech giants ban micro-targeting: important?

Over the past few days there have been credible reports (see this Guardian piece) that both Google and Facebook are seriously considering a ban on micro-targeted political ads. Tech giants have a lot of thoughts they don’t follow through but if this happened it would be good news for everyone because those ads are total garbage. Such a ban would have a considerable impact on American PACs but won’t make much difference to UK campaigns. Here’s why.

As a starter, let’s explain what the platforms are talking about. Micro-targeting involves specifying recipients of your campaign based on their freely shared demographic and life stage data combined with their location, to a level that makes it statistically probable you are targeting a particular household or block – and adjusting your content accordingly. Microtargeting is intrusive and presents particular ethical quandaries in a political context, with many concerned about the prospect of different groups being completely siloed in terms of the messaging they receive. A letter from Facebook employees to Mark Zuckerberg detailed practices in which political advertisers would use the electoral roll with the Custom Ads audience to develop an audience of voters, who they would further track using offsite behaviour trackers and ad engagements. Essentially these members of the public are being submitted to customised, intrusive advertising at an unprecedented level with little ability to understand where it is coming from or how far to trust it.

The ability to micro-target using both Facebook and Google ad platforms is what enabled charlatans like Cambridge Analytica to claim they could affect voter turnout and win elections, when all they had in their back pocket was some basic profiles built on second hand quiz data. And it is a matter of record that Bannon’s Trump 2016 network made significant use of micro-targeting in the 2016 US election to depress turnout among black voters and stoke fear among white voters. Micro-targeting has been used in Europe and the UK to ensure only susceptible people see misleading or inflammatory ads. Such material is usually pushed out not by campaigns but by their clandestine surrogates.

So it is easy to see why the platforms would want to act against micro-targeting. It is really bad (and contrary to the ideals of Silicon Valley) to see your tools being used to sew division and to reduce participation in democracy. Both Google and Facebook have taken steps to let people see how political ads are being targeted, and in our work at Signify we’ve made use of their political ads libraries in the past, but it is evidently unrealistic to expect individual voters to investigate the massive quantity of different messages being put out by parties to people matching different targeting criteria.

However, micro-targeting is also niche and not used much by mainstream political campaigns who are trying to mobilise entire constituencies. In the UK, the main parties have already created their own governance filters that prevent this level of targeting such as the Labour ‘Promote’ tool. In terms of online campaigning the most effective teams set demographic constraints but keep these at a district-level in terms of location. For instance, the winning Vote Leave campaign in 2016 simply avoided Remain towns. That isn't micro-targeting.

There are also many more impactful things that the platforms could do. Such as banning or sanctioning politicians who promote lies. Which Facebook is not going to do, despite some pretty harsh mainstream criticism and calls for regulation. You could even argue that publicly acting on micro-targeting would be a diversion from more serious regulation.

Zuck meets the press

But with all that said banning micro-targeting would be a step in the right direction. It is gross. And campaigns (both political and brand-led) need to step away from the creepy and hyper-targeted model of social engagement. They must stop being intrusive and start being attractive.

There is a better model for the use of social data, espoused by companies like Signify and by many progressive campaign organisations. That model is to find the things that thousands or millions of individuals care about – like healthcare – and to build paid and organic campaigns that appeal to people from all walks of life who care about their livelihoods, their society and the future of their children. Of course, these campaigns use filters to avoid campaigning in areas that they cannot win, but they don’t need to think about stopping turnout street by street or person by person, because they are built on positive action. The more people see and engage, the better. This is a structural advantage for campaigns built on truthful responses to issues that matter.

Which brings me nicely to the current UK campaign. All the major parties will be aggressively targeting each other’s voters, but micro-targeting is already ruled out by social media governance agreed by all parties and enshrined in their actual ad software. A caveat to that is the Brexit party, who are still only a limited company and have shown little regard for regulation in the past.

Specifically, the one campaign organisation and persona that you might think would love something as creepy and intrusive as micro-targeting – Dominic Cummings and his surrogates around Westminster – will likely make no use of micro-targeting tools. Their strategy is based on a public, futile battle (People vs Parliament) offset by a much more effective but hopefully losing recruitment strategy that lures voters away from Labour and the Liberal Democrats or solidifies their Conservative vote all based on misinformation and fear. The recruitment part of their campaign is categorically not based on microtargeting but on a spray gun attempt to confuse millions of undecided or wavering voters.

Cummings: not a fan of micro-targeting

The parties running against this Conservative government have a key structural advantage which is the ability to tell the truth about the last nine years and what needs to change. Signify are supporting local and national efforts to raise awareness around the issues that matter to voters. What we will never do is use micro-targeting to deliver messages. Why? Firstly, because it is creepy and wrong. Secondly, because everything our contacts are saying is true and we don’t care who knows it. There is no need to say one thing to one group, and something different to another.

Therein lies the real ethical problem at the heart of micro-targeting: if you are saying something to one person, but don’t want anyone else to even see it – why not? 

To find out more about Signify please get in touch.