Three revenue lines, in the order they come true. The arithmetic is here so you can redo it, and the part we have not proved yet is named in the same breath as the part we have.
Two of these are software subscriptions with ordinary economics. The third only exists because of what the first two build, which is a network of people who reliably turn up.
Any product whose users would be better off knowing each other. One embed, every city, unlimited users. Under a grand on purpose: it fits a manager’s discretionary spend, so the person who wants it can buy it without procurement. The same job costs $8,000 to $33,000 a month as a hire.
A thousand customers is a real sales motion, not a viral one. This line pays for the network, it does not by itself get us to a hundred million.
Member bases in the hundreds of thousands to millions. Charging a company with five million members $999 is not a discount, it signals a toy, and it loses money on curation at that scale. Price on how many people we program for.
These are six to nine month sales cycles with security reviews. Which is why the SDK ships pinned versions, a documented CSP and a browser test that runs cross origin.
Once the network reliably puts thousands of people in real rooms every week, being there is worth money to a brand, and worth more than an impression because someone showed up. This is what takes the company past software subscriptions.
We will not sell this until attendance is real and repeatable. Selling it early would mean promising rooms we cannot fill, which kills the other two lines.
Every step is in production today. The first thousand customers should never speak to anyone here.
In the order the company needs them. Bands and equity are on the careers page.
Not our numbers, and we will not pretend they are. Published studies, and the reason this budget exists at all.
SaaS customers active in a product community, versus inactive ones.
Gainsight, 2023
Community members compared with users who never joined one.
Orbit Media
Still holds after controlling for how much they use the product, so it is not just that engaged people join things.
Community industry research
Members answer each other. Companies report saving an average of $145,000 a year on support alone.
Community industry research
Two caveats. That research is mostly about online communities. Ours is offline, which we think is stronger and cannot prove yet. And none of it is ours: no customers, so no retention curve. What we built is the instrument that makes one. Every gathering records who committed and who arrived.
The arithmetic, without us: 100,000 users at $10 a month lose about $1.2M a year per point of monthly churn. Soul is $11,988 a year. It has to move retention by a rounding error to pay for itself. If it moves nothing, the brief says so.
Not a theory about whether people leave the house. Coffee shops, rooftops and bars, offline, before a line of this product existed.
Built and run by us, before there was any software to do it with.
More than three a week, every week, programmed and filled by one person. That workload is the entire reason this product exists.
Per gathering. Not an invite list and not a signup count, the number who were actually there.
Three a week, every week: choosing it, writing it, filling it, standing there. One city, one person, no staff. That is the ceiling of doing this by hand, and exactly the ceiling the agent removes.
There are no photos of any of it. Nothing posted, nothing streamed, which was the rule rather than an oversight, and the same rule the product runs on. The record we keep now is who turned up.
Including the one we cannot answer yet.
They already did. 10,000 people, 100 gatherings in six months, 80 to 90 in the room, offline, by hand. What is unproven is an agent doing it for somebody else’s users in a city we do not live in. That is what the next ten gatherings are for.
Both wait for somebody to have an idea. Soul decides what to run, every week, unasked, inside their product. We replace the hire, not the tool.
They can run a gathering for their own users. They cannot start with a network that already holds other companies’ people in the same city. In house means an empty room in every market at once, and an empty room is not a cheaper full one.
Nothing, and we do not claim otherwise. No patents, none pending. What cannot be copied is the record of who turned up, which only exists because it already happened, and the position across many companies that lets one gathering belong to two brands.
The product is live: partner boards, a pinned SDK, real OAuth, a payment path that mints a key. No paying logos and no attendance record yet. We would rather say it than dress a demo as revenue.
An agent can now program something specific and good for a group, weekly, in a city it has never seen. That used to need a person on payroll in every market. And loneliness stopped being soft and became retention.
The network compounds across customers, not within one. Every company that installs Soul makes the matching better for everyone else’s users in that city. No in-house tool gets that.
Not features, not logos, not a deck. Ten gatherings in New York with an honest count of who came and who came back. Everything on this page is either already running or waiting on that number, and we would rather tell you which is which.
Talk to us →