What Cohort 0 Can Prove — and What It Cannot

Cohort 0 is the first group of members, merchants and agents to go through the full model. Its results will be published. What it can and cannot demonstrate is worth agreeing before anyone sees a number.

What Cohort 0 can settle

• Whether the mechanics work — escrow funding, application, selection, confirmation and payment, end to end
• Whether merchants post work once they are past 14 days and 50 sales
• Whether members earn anything at all, and roughly how long a first win takes
• Whether agents survive to month six
• Where the process breaks — the steps people abandon
• Whether free users convert, on either side of the market

These are operational questions, and a first cohort answers them well.

What it cannot settle

Whether the model scales. A small cohort receives attention that ten thousand members cannot. Community micro-tasks that are "always available" to a hundred people may not be to ten thousand — that is the largest open risk in the model and Cohort 0 cannot test it.

What typical outcomes look like. Early participants are self-selected, unusually motivated and unusually supported. Their results describe them.

Long-run retention. Six months of data cannot tell you about year three.

Whether merchants fund the second round. During launch Knomart funds diamond redemption and refunds merchants. Whether merchants pay for it themselves afterwards is the business model in one question, and it sits beyond this cohort.

Why publish the limits first

Because a result published without its limits gets used to prove more than it can. Stating in advance which questions a cohort answers is what stops a modest operational finding being presented as evidence the model works at scale.

What will be published

The measures already named — member economic value, conversion on both sides, earning rate, applications before a first win, retention and agent survival — with their real values, and with this list of limitations attached.