For the past few months I've been building Socion, a matching app built entirely around socionics intertype relations.
It went live this week at socion.app.
The idea
Most matching products are black boxes. A private algorithm optimises for engagement, not compatibility, and you never know why you're being shown someone.
Socionics is different. It maps 16 named relationship dynamics between every possible pair of types: not just compatible or incompatible, but characterised relationships with predictable patterns. Duals complement each other naturally. Mirrors stimulate each other but tend to criticise. Conflict drains both sides regardless of effort or goodwill.
Socion puts that theory directly in users' hands. You pick your type, choose which dynamics you want to match on, and browse profiles whose type produces that relation with yours. The matching logic is the intertype matrix: open source and auditable.
Why it's interesting as a data problem
Socionics makes predictions that can be tested. If the theory is right, Dual pairs should report better connections than Conflict pairs at scale. Mirror pairs should produce more messages. Certain dynamics should go with longer conversations.
As far as I know, no one has tested this at any meaningful scale in the English-speaking world. The app is designed to generate that data, through post-match ratings, conversation length and connection outcomes, and find out whether the theory actually holds.
That's the product analytics angle that interests me most. Not just building a matching product, but building one that can confirm or challenge the theory it's built on.
What's live
- Type onboarding (questionnaire, or bring your own type)
- A matching feed filtered by intertype relation
- Messaging with real-time delivery
- Post-match feedback
- Open source code at github.com/sstern42/socionics-match
If you know your socionics type and want to try it, it's free at socion.app.
And if you're a researcher interested in personality type and relationship dynamics, I'd welcome a conversation about what this data could produce.