How your Circle gets made: the archetypes and the science

People ask what happens between finishing the questions and hearing "we found your people." Here's the honest answer, research and all.

Most matching products treat their method like a trade secret. I think the opposite is true for friendship: if we're going to introduce you to five strangers and suggest you spend your Saturday with them, you deserve to know exactly why these five. So here is how a Circle gets made, start to finish.

Step one: a set of honest questions

It is 28 questions. There are no trick questions and no right answers. It measures a handful of things that decades of personality research say actually matter for friendship: where your energy comes from, whether you plan or drift, how hungry you are for new places and ideas, how warm you run, how much you value honesty over polish, what your humor sounds like, and whether you'd rather have one real conversation or a night of easy banter.

It also asks about your real life, because compatibility on paper dies on a calendar. When are you actually free. What a hang should cost. Whether plans should involve a bar, either, or neither. Which part of your city you're in, and how long you've been here. That last one matters more than people expect: two people who both landed in the same place eighteen months ago already share something worth building on.

Step two: your archetype

Three of those measures combine into your friendship archetype: your social energy (steady or vibrant), your rhythm (planner or free-spirit), and your compass (rooted in the familiar or pulled toward the new). Two settings on three dials makes eight types.

The labels are the headline, not the whole engine. Underneath, the same answers feed a compatibility model that also weighs warmth, steadiness, honesty, humor style, conversation depth, budget, and schedule. But the archetype earns its place, because Circles are built on it in a way that might surprise you.

Step three: a group, mixed on purpose

Here's the part most people get backwards. A good friend group is not five copies of the same person. Put five quiet observers at a table and everyone waits politely for a conversation that never starts. Put five conversation-drivers together and they spend the night interrupting each other. Research on small groups keeps landing in the same place: mixed groups work better, especially once people are interacting often.

So the engine composes each Circle like a seating chart. Everyone shares one anchor interest, the thing that gives the first hour its subject, because half a century of friendship research says similarity is where friendship starts. Then it deliberately mixes the rest: a balance of steady and vibrant energy, at least one natural organizer so plans actually happen, and, when someone in the group finds new rooms nerve-wracking, someone whose instinct is to pull quieter people in. Everyone overlaps on at least two windows of the week when they're genuinely free, because a Circle that can't schedule is a group chat that dies in March.

No one hand-picks your table. The engine composes each Circle and checks it against its own quality bar before it goes out. The design is deliberate: people, not staff, run their own Circles. The only time a person steps in is enforcement, if something is reported, a human reviews that.

Step four: verified, then revealed

When your Circle is ready you see first names, archetypes, a shared interest, and neighborhoods. Not a stack of photos to judge. People are previews here, not products. To take your seat you verify you're a real person: free, once, about two minutes, handled by our verification partner rather than stored by us. Showing up to a patio table with five strangers takes a little courage. Knowing every one of them is real is what makes the courage reasonable. (Nervous about what to actually say once you're there? We covered what to talk about at your first Circle meeting.)

Step five: the 200 hours are yours

Jeffrey Hall, a communication researcher at the University of Kansas, put numbers on something everyone suspected: moving from stranger to casual friend takes roughly 50 hours together, and close friendship arrives somewhere past 200. Crucially, the hours that count are chosen, unhurried ones. Time at trivia counts. Time in meetings doesn't.

That finding shaped the product more than any other. It's why a Circle isn't a one-off dinner but a group with tools to keep going: an availability poll, venue ideas near you, one tap to set the next meetup. It's why the icebreaker card at your first meetup is light, and the ones that come later go a little deeper, following what psychologists have learned about how gradually opening up turns acquaintances into friends. The platform can't manufacture your 200 hours. It can make sure none of them are wasted on the wrong table. (More on the mechanics: how to make friends as an adult and why a Circle is four to six people.)

What the system won't do

It won't change your profile behind your back. If the pattern of your RSVPs suggests weekend mornings suit you better, it asks, and you tap keep or not right. It won't show your Circle anything about your dating life, ever, including whether you have one. And it won't pretend an algorithm can guarantee friendship. Nobody can promise chemistry. What we can do is remove the bad odds: the stranger with nothing in common, the audition dinner for two (no friendship application required, we already did that vetting for you), the group that never finds a date that works. The rest, happily, is up to you.

For the full citations behind this, Dunbar's group-size research, the 50/90/200-hour study, and where the evidence is genuinely mixed, see the science behind how Circles work.