Matchmaking & Relationship Services
Matchmaking is a barbell with an empty middle. Apps are paid for engagement; bespoke matchmakers cannot scale a hand-worked search. The middle is now buildable: machinery carries the search, a thin human layer keeps the judgment.
The product in this category is not introductions; it is trust — verified identity, screened intent, consent handled with dignity. The operator who can compute what neither pole can, on verified data neither pole collects, owns the price point between them.
Industry research
The economics
FIGURES FROM THE FIRM’S RESEARCH · SOURCES ON REQUEST
Growth opportunities
Where the 2× lives
The search, industrialized
Intake, profiling, sourcing, and compatibility ranking run as one pipeline — in the firm’s build, five scoring layers blending three embedding channels — with the human concierge kept for the two calls that matter: the chemistry read and the deception catch.
Trust as the product
Verification fail-closed at the database layer, staged so it never kills the funnel: light screening at signup, the full identity ladder fired only when an introduction is actually in prospect. Verified affinity becomes a matching signal competitors cannot compute — because they do not verify.
Dignity as architecture
Consent and decline handled structurally: a woman-first introduction sequence in which the paying client never sees a decline, enforced by a shared state machine rather than concierge tact. People join a market where declining is safe. That is the whole recruiting strategy.
Regulation as a design input
State dating-service law — price caps, cooling-off rights, term limits — carried as data in the compliance layer, statutes cited in the code comments, checkout gates defaulting closed pending counsel sign-off. In the firm’s build this lives in the schema.
Our perspective
What most firms get wrong
Competing with the apps on their field
The instinct is to build a better swipe. But the apps are not failing at their job — their job is engagement, and they are excellent at it. The open position is the one both poles concede: outcomes, verified and priced accordingly.
Where we’d start
A search-labor decomposition: hours per completed introduction split into sourcing, screening, and judgment; the automatable share priced; and the verification ladder plus the state-law constraints for every operating state, specified before the first introduction.
Put it to us