For investors · Updated 14 August 2026
Meeting someone used to happen while you were doing something else.
The record follows01 What this is
Heartstrings is a connection company. The dating industry made the introduction efficient and left what happens after it untouched, which is where most connections die.
We build co-operative games that two people play together, and the company earns from the game library rather than from keeping anyone single.
02 What exists
Heartstrings has run as a live platform for thirteen months. It is in production today, on Discord, operated by one person. Every profile on it passed a thirty-second live voice call with a human before it was approved. It was built and operated with no funding and no paid acquisition.
There is no company entity yet and no standalone app yet. The figures below come from the platform that is running.
03 The record
- 6,168
Audited join events
Read directly from the server invite panel on 1 August 2026, across 39 active links. A floor: expired and deleted links are not listed and their uses cannot be recovered.
- 31.8%
Passed the verification gate
1,962 of the 6,168 audited join events passed the live voice call. This is the gate on its own. It is not multiplied by anything that happened afterwards.
- 614
Profiles surviving
1,962 passed the gate. 614 hold a profile today. The rest have departed, and the departure is treated separately in Section 04.
- 983
Lifetime matches
Reconstructed from the record of notifications the platform itself sent, not from anyone's conversations. It can only undercount, so it is a floor.
Both people in a match are notified, so one event writes two mirrored records. The archive's 1,522 records collapse to 1,115 notifications across 983 distinct matches, and the 407 collapsed duplicates are reported rather than quietly dropped. All of them authenticate against the messaging platform's own message identifier, 1,522 of 1,522.
- $1,345.15
Lifetime revenue, net
$1,561.65 gross across thirteen months to 26 July 2026, through a third-party paywall. 14 accounts are paying at that export, $20.73 per payer: one account is $150 of it, the other thirteen average $10.79.
Three capture dates. They are not reconciled to one another and are not presented as though they were. Nothing above describes today.
04 How we handle our own numbers
Corrections run in whichever direction the data goes.
Every figure this company publishes carries its denominator and its as-of date. When one turns out to be wrong it is corrected in public, with the old value printed beside the new one and the reason attached. Three that ran against us:
97.4% of users ever matched
68.5%, being 318 of the 464 who liked someone
The old figure divided 598 lifetime identifiers by the 614 current profiles, when 1,962 accounts had passed the gate across the platform's life. 273 of those identifiers belonged to people who had already departed. Overstated by 45 points.
25.1% all-time patron churn
74.0%, being 54 of the 73 who ever paid
The old denominator counted every roster row, including 146 followers who never pledged and therefore could never churn. This is the company's headline monetisation risk and it is three times worse than first reported.
Roughly 7,000 views and 2,300 reads on our flagship article
8.8K views and 2.7K reads, captured 27 August 2026
The original claim described the 23 June 2025 flagship article and was accurate and conservative. A prior correction compared it to a different article, published 8 April 2026, whose statistics showed 149 views. That correction was withdrawn after the correct article's statistics page was captured at 303 presentations, 8.8K views and 2.7K reads.
One more, printed because a reader would otherwise do the subtraction.
1,962 people passed the voice gate across the platform's life and 614 hold a profile today, so at least 1,348 have left. That is a survival ceiling of 31.3% and a departure floor of 68.7%.
05 Where this is going
Stated as a bet, because that is what it is.
A library of co-operative games sits in front of the connection layer, released in seasons. The wager is that an experience worth sharing acquires people below dating-app cost, where another profile to swipe does not, and that a library gives a reason to renew that a matching service structurally cannot give. Revenue then comes from the games rather than from anyone staying single, so a user who finds somebody is a customer who keeps paying instead of a subscription that ends.
What is not known, said before it is asked. Paid acquisition has never been run here, so the cost of acquiring a user is unmeasured rather than modelled. The automated verification that replaces the voice call has no activation rate yet. Nothing in the record above shows whether a match worked, because the Discord schema logs no match outcome at all. Those three are what the money buys first.
06 The round
Before any round opened, supporters were already paying between $8 and $150 a month on the vision alone, for a studio that has not shipped a game. Nobody was asked.
- Team45%
- Product and the first title30%
- Growth, the first real acquisition test15%
- Operations10%
07 Who to write to
For anything else: heartstrings@zanginc.net
運命の赤い糸
Unmei no akai ito
The thread may stretch.
It may tangle.
It never breaks.
Unmei no akai ito, the red thread of fate, is the belief the company is named after. It shapes what gets built and how this page is drawn. It is not part of the case above, and it is printed here rather than folded into one.