Platforms and algorithms · intermediate

Microtargeting

The message found you because it was built for you. It was never meant to be seen by anyone else.

Also known as Cambridge Analytica-style political microtargeting

Persuasive messages tailored to a narrow audience using data about that audience's fears, habits, and vulnerabilities, then delivered where only that audience will see it. The same campaign can tell one neighborhood the candidate will protect them and tell another that the candidate will punish them, and neither group ever sees the other's version. The power is not in any single lie. It is in the invisibility: because the message is private, it is never publicly tested, never fact-checked against its sibling messages, and never held to account for contradicting them.

Truth-adjacency

Truth-adjacent: the pattern's significance depends on whether the claim is true

Where it shows up

Platforms and algorithms

What to watch for

The phrases and tells that mark this pattern in the wild:

an ad that feels unnervingly specific to your situationdifferent people seeing different versions of the same campaigna promise made in a forum no journalist monitorsmessaging that references your recent search or locationa candidate's position that seems to shift depending on who is asking

How to recognize it

The tell is the contradiction you cannot see from inside your own feed. Microtargeting is invisible to its targets by design: each audience receives a message built for them and assumes it is the whole message. The pattern surfaces only in comparison, when two people discover they were shown opposite versions of the same campaign. Watch for messages that feel built from your private data, promises made only in narrow channels, and the absence of a single, public, testable position. If a speaker’s stance changes depending on which room they are in, and the rooms are chosen so they never meet, that is the architecture of the pattern.

What to ask

What it looks like when you’re wrong about it

You call “microtargeting” on a campaign that simply runs different ads for different audiences in good faith, with messages that are consistent in substance and could all be shown together without contradiction. Audience segmentation is normal and often useful. The pattern requires the invisibility to be load-bearing: the audiences are kept apart so contradictory promises are never compared, the targeting exploits inferred vulnerabilities, and the messages are designed to evade public testing. If the different messages could sit side by side in daylight without embarrassing the sender, you are looking at segmentation, not the pattern.

What it feels like from the inside

How it starts

A campaign uses behavioral data to identify audience segments and crafts messages specific to each group's profile, delivering them through private ad channels rather than public forums.

How it progresses

  1. Messages become increasingly specific, referencing recent searches, location data, or inferred psychological traits the person did not knowingly provide.
  2. Contradictory promises circulate simultaneously: one audience hears protection, another hears punishment, neither knows the other exists.
  3. The sender never commits to a single, public, testable position. Their stance changes depending on which room they are in.
  4. Public accountability becomes structurally impossible because no single audience sees the whole picture, and the whole picture is where the contradictions live.

Common signs

Why it's hard to leave

The message was built for you. It addresses your specific fear or desire, which makes it feel relevant rather than manipulative. Rejecting it means rejecting something that seems to understand you.

Do this now

  1. Compare notes. Ask people in different demographics what they were told by the same source. Contradictory private messages are the signature.
  2. Ask: could this message survive being shown to everyone? If the sender would be embarrassed for other audiences to see it, the privacy is the strategy.
  3. Check what data made you a target. If the targeting depends on inferences you did not knowingly provide, the message is built on surveillance, not relevance.

What people realize later

Later, people realize the message was not a conversation. It was a product of surveillance, tailored to a profile they did not build, delivered in a channel designed to prevent comparison with what other audiences received.

Recognized this online?

Misuse Guardrails

How this pattern gets misused

Someone treats any personalized advertising as sinister microtargeting, including benign uses like showing a local restaurant ad to people nearby. The term becomes a way to frame all data-driven marketing as manipulation, which makes it harder to focus on the cases where the tailoring is genuinely exploitative, such as targeting a vulnerability the data has revealed.

What it looks like when you're wrong about it

Tailoring a message to a relevant audience is ordinary communication, not microtargeting in the manipulative sense. A charity showing animal-rescue ads to people who follow animal-rescue accounts is matching interest to content. The pattern requires the tailoring to exploit a revealed vulnerability and to depend on invisibility: different audiences get contradictory messages, the targeting uses sensitive inferences the person did not knowingly provide, and the messages are designed to escape public scrutiny. Relevant targeting that could survive daylight is marketing. Contradictory targeting that requires the dark is the pattern.

Not sure? Describe the situation to someone outside it. If they do not see the pattern, pause before you name it.

Related Patterns

The name is designed to spread. The hook is designed to stick. If you recognized something, share the name.