Most data deals are one-off sales of a dump. We push for licenses with a refresh term instead, so the same dataset keeps paying as your company keeps working.
The history sells first
The first license on a dataset covers its history, every past record the source holds up to the day of the first run, and has its own price for that history.
New records go out every month
After the first run, the SDK runs on a schedule, monthly by default. Each run pulls only the records created or changed since the last successful run, scrubs them on your machine and uploads them as a new batch for that month. A batch covers one dataset and one period, written as YYYY-MM.
Each refresh term pays per batch
A license can include a refresh term with a price per monthly refresh. Every batch that is accepted pays that refresh price once for each license with an active refresh term. Two licenses with refresh terms on the same dataset mean two refresh payments for each accepted batch.
| Event | What pays |
|---|---|
| First license signed | The license's price for the history |
| Monthly batch accepted | The refresh price, once for each license with an active refresh term |
More work means bigger batches
Batches grow with activity. A team that sends more invoices, closes more deals or handles more tickets creates more records each month, so each refresh batch carries more data. As the team grows, so does the value of each refresh.
No work after setup
Once the SDK is scheduled, there is nothing to do each month. The scheduled run pulls, scrubs and uploads on its own, and the portal shows each batch as it moves from received to accepted. You come back when we ask you to approve a new buyer, or when something in the troubleshooting list needs fixing, such as an expired token.
Your share
Each payout records your share in basis points. Your seller agreement sets the share. See Payouts.