Yesterday, Ursula Von der Leyen announced that the EU is creating a new unit to counter disinformation.

Disinfo and populism are tearing the EU apart, but it's not just because trolls and populists leverage bald-faced lies. The facts aren't actually what's most important here. What's important is what people believe. And the reason the trolls are winning is because they tell stories that work for the segments they're targeting.

Often, those stories paint institutions like the EU as the cause of their targets' pain. Every story needs a good bad guy, after all. And as trust in the trolls' narratives grows, trust in institutions declines.

And so if the trolls' win by undermining trust, then the best way to respond isn't with fact checking. Fact checking doesn't work (not least because the only people motivated to engage with facts are the people who value facts already). The best place to start is by strengthening trust in institutional brands.

Charlelie Jourdan makes the point that Brexit didn't start in 2016. It began when the EU never bothered to show Brits how it was a win for them. The areas of Britain with the highest levels of EU funding were the places that voted hardest for Leave. If an institution doesn't tell a compelling story / brand, it leaves itself as an open target.

So this digs into why the EU, the UN, most centrists and most NGOs fail so spectacularly at building mainstream affinity with their brands. 

And probably, first and foremost, it's a philosophical problem. All the people that run government and NGOs, the EU and the UN are enlightenment thinkers. They love ideas, and they believe the best ideas will win in the end.

The thing is that that's not compatible with how most people work. Almost everyone else is a romantic. We’re driven by our emotions, by our perceptions of in-group and out-group, culture, belonging. We’re not motivated to seek truth. We’re motivated to get on in the world, to find our place and prosper. Sometimes truth is useful for that. A lot of the time it’s not. Take religion, which is probably the most successful example of an ideological meme.

I don’t think we need to discuss the factual validity of Jesus’ ascendance to heaven on the third day. Instead, what’s far more interesting to look at is the trajectory of the various religious memes over the millenia in how they perpetuated or died out according to the opportunities they afforded their followers for social currency, grouping and power. 

Religion was, in many ways, the first mass media. Churches and temples pushed out on-brand messaging in every village for a thousand years.

Now, media is digital. And it is very much the catalyst for democracy’s decline. And it’s worth looking at how philosophy determines our approach to media creation. According to enlightenment philosophy, if we put out an idea, that’s our job done. The idea will win in the end if it’s good enough (and our ideas are the best, of course). That makes it very difficult for enlightenment thinkers to understand how new media works.

Enlightenment thinkers enjoyed a couple of hundred years where their ideas dominated. And that dominance entrenched the belief that ideas win on their strength alone. But actually, between the invention of the printing press and the advent of "The Facebook", media was enormously expensive to produce and distribute. That meant it was scarce.

So for a while it looked like the enlightenment ideal was working out really rather well. In fact, what was happening was that consumers didn’t have much of a choice. So we had an illusion of consensus because a few enlightened thinkers got to decide which ideas got traction. 

Now, anyone can put out anything, and anyone can engage with anything. And no one voluntarily engages with stuff that doesn’t interest them. (Quite obviously) 

Algorithms aren’t designed to spread hate or to suppress democratic ideas (as many in the enlightenment community mistakenly believe). They’re designed to optimise for clicks. It works like this: if more people than average watch a video through to the end, the algorithm will categorise the kind of people who are watching, and it will show it to more people whose previous online behaviour puts them in the same category.

Statistically speaking (and algorithms are statistical models) anyone in the same category is more likely to watch to the end too. (Of course, emotion is what drives those clicks, but the emotional tone of the videos is not how the algorithms determine how far they’ll push them).   

It’s a system designed to keep people online as long as possible. And the purpose of keeping them online longer is to connect them with brands that want to sell them stuff. Your media consumption habits are a great proxy for your social class, income and consumption habits.

And your online history primes the algorithm to serve you relevant ads. And because everyone has different interests and different values, it shows you what you’re interested in, while in the same moment, it shows your neighbour something completely different. It’s a system designed to please all of the people all of the time. 

Soon, Meta and Google will discontinue paid political advertising in the EU. This will be another nail in the EU / democratic / INGO coffin. It will make polarisation much, much worse. And it’s a huge win for the extremists and the populists - who are very much Romantics in their approach - and who are already doing splendidly with their organic engagement.

The EU will still be able to post organically, of course, but all organic engagement is echo chamber engagement. So, for example, if the EU puts something out that’s supposed to combat disinformation, then the only people who will be served that content are the kind of people who are already pro-EU and motivated to engage with facts. And only reaching people who know the facts already achieves precisely nothing. It just entrenches polarisation. 

So, what do we do?

Lots of other media thinkers are writing about using earned media, organic, influencers etc to replace paid ads.

But before they dig into tactics, institutions need to redefine their objectives. Their objective needs to be to build affinity with democracy and facts with the maximum possible audience of persuadables. Not the people who are with them already, and not the people who are always going to hate them. They need to build affinity with the vast majority of people who think politics is boring, and who, because of the system I just described, are gradually losing touch with democratic ideas and facts. 

This is a question of brand building vs performance marketing. It's a big discussion happening in the advertising community at the moment. Basically, it's about getting instant sales from a small community of fans vs building affinity with your story (and a larger eventual volume of sales) with a much bigger community whose clicks you can’t directly track. Check out Tom Goodwin or Byron Sharp for smart stuff on this if you’re interested (and if you work in political media YOU SHOULD BE). Broadly the conclusion is that algorithms are shrinking sales for big brands.  

There are direct parallels with politics. Unfortunately most people in political and NGO comms won’t know what brand building and performance marketing mean. And that knowledge gap means that every NGO and political institution has only been doing performance marketing since social media began. They design their posts to get organic engagement from people who are already with them.

Likes are how comms teams are incentivised, and the easiest people to get likes from are the people that like you already. And if you design your content for people that are with you already, that means you're using ingroup language and ideas that make it less accessible to anyone else. And for everyone else, well, in a system designed to please all of the people all of the time, there’s always something else on offer. 

The starting point of effective messaging is to root strategy in a solid understanding of your target audience’s values, interests, culture, beliefs, watch habits, etc. That is just not happening in most places doing political messaging. (But it is happening on the populist side and in the troll farms in Moscow, where they have very smart strategists doing quantifiably effective media). And that means that only the values and culture of the people doing the EU's messaging tend to be reflected in the EU's messages.

That is, middle class, well educated, doing pretty well, thanks very much. And again, the enlightenment perspective gets in the way of them understanding other people’s motivations to engage. They write for themselves and their community, in other words. And in a system designed to keep all the people happy all the time, that’s what's got us where we are today. 

Most people in institutional comms think their job is: say what we need to say in the best way we can say it. What their job descriptions should actually be is this: say what we need to say in a way that ensures the macro optimal audience chooses to engage with it.

With a bit of basic research, we can quantify our macro optimal audience, and suddenly we have a measurable objective. And then if we apply techniques that are standard in the private sector now - experimentation, data analysis and iteration - we can make sure our traction grows with the segments that count. And if we can start building traction, we can start building the brand affinity the EU needs.