Social media platforms were built as advertising machines. Their core business is to connect a vegan smoothie brand to health‑obsessed urbanites, or a cheesy sunglasses label to the subcultures that love ironic fashion. They do this by clustering people into behavioural niches, then selling advertisers access to those most likely to respond. That same machinery has, almost by osmosis, reshaped political campaigning.

Civil society organisations plug their content into automated targeting tools and “lookalike audiences” that faithfully find more of the people who already agree. A climate NGO’s AI‑generated ad and its human‑made fundraising creative look almost identical because both are optimised for the same outcome: the cheapest engagement from the most committed believers.

When a European institution celebrates a video with 14,000 likes, it often turns out those likes came from loyal fans who needed no convincing at all.

Echo chambers as a rational but self‑defeating strategy

From the inside, this behaviour feels sensible. NGOs need donations to survive, parties need visible enthusiasm, and activists crave signs that their messages are landing. Automated targeting delivers all three with impressive dashboards attached. Lookalike tools offer a low‑friction way to find people who will sign the petition, attend the rally, or retweet the campaign.

The problem is what this does to the broader information environment. As democratic actors sink more of their budget into audiences who already agree, their ideas become less visible in the everyday media diets of the uncommitted majority. In that vacuum, well‑designed disinformation and grievance politics find fertile ground. The likes tick up on progressive pages, but outside those bubbles, the conversation is shifting.

How bad actors mastered the mainstream

Populist strategists and well‑funded PR firms play a different game. They start from the concerns of ordinary, non‑political people: rising prices, insecure work, a sense that politics is remote or contemptuous. Then they iterate narratives until they find framings that cut across ideologies.

The line that “net zero is a tax on the poor” did not arrive by accident; it was honed in research labs until it resonated with broad swathes of mainstream audiences. Once the data showed it struck a nerve, it became the cornerstone of a campaign that moved climate politics from “is it real?” to “can we afford it?”

That is what it looks like when political communication is treated as a precision business: define a maximum persuadable audience, test messages, and invest in the ones that move people who are not already on your side.

Re‑pricing engagement: targeting the medium‑cost middle

If democracy is to stop losing this math problem, it must change what it optimises for. Instead of cheering low‑cost engagement from ideological super‑fans, campaigns need to reward medium‑cost engagement from the persuadable middle. At one end of the spectrum sit citizens who will never be convinced; at the other are those who are already die‑hard supporters. The leverage lies in people close enough to listen but not yet committed.

That means adopting the tools of modern marketing, but with different values baked into the brief: define persuadable audiences, run systematic A/B and multivariate tests, and compare the cost per meaningful interaction across mainstream and base segments. When NGOs have taken this approach, re‑cutting familiar campaign videos into fresher, less insidery formats, they have seen double‑digit improvements in mainstream engagement without spending more money. The mainstream is not unreachable; it is just bored of being preached to in a language that signals “this is for activists.”

Choosing discomfort over cheap affirmation

Regulation is making parts of this work harder, Europe’s new rules on political advertising have tightened the screws on targeted ads, but the underlying choice remains. Democratic actors can keep chasing the instant gratification of base applause, or they can accept the discomfort of speaking to people who do not yet like them. The former keeps organisations afloat in the short term; the latter keeps democracy afloat in the long term.

Funders and regulators have a role to play in rewiring incentives: tying grants and evaluations to demonstrated impact beyond the core base, and encouraging experiments that measure shifts in mainstream attitudes rather than total impressions.

Ultimately, the question for pro‑democracy forces is whether they are prepared to pay the true price of persuasion, time, creativity, and budgets aimed at the middle, rather than the discounted comfort of talking only to themselves.