Variety of cases
The more varied your data, the more useful it is. A dataset with 200 different presentations of the same condition tells us more than 2,000 nearly identical ones. We're trying to study the full range of real cases, not just the common ones. Variety is the single biggest factor in what we pay.
Number of distinct procedures or experiments
Institutions that cover a wider range of procedures, conditions, or experimental setups land toward the higher end of the range. A pathology lab that handles fifteen different cancer types provides more research surface than one focused on three. A sequencing lab that runs five different assay types is more valuable than one that runs one.
How thoroughly each case is documented
Raw results alone are less valuable than results with the practitioner's reasoning attached. If a case includes what was being looked for, what alternatives were considered, and why a particular conclusion was reached, that's much more useful to us than the same case stripped down to just an outcome. Documentation depth often moves the price more than raw volume.
Rare and unusual cases
Rare cases are the hardest to study from existing public datasets, which makes them disproportionately valuable. If your institution sees presentations or experiments that aren't easily found elsewhere, such as unusual conditions, atypical responses, or novel procedures, that pushes the engagement toward the upper end of the range.
Volume
All else equal, more cases is more value. But variety almost always beats volume: 500 well-documented, varied cases is more useful than 5,000 nearly identical ones. We'd rather have a smaller dataset that covers more ground than a large one that doesn't.
Anonymization effort
The work of anonymization is part of what we pay for. If your existing pipelines already strip identifiers cleanly, the engagement runs more smoothly. If we need to work with you to figure out how to safely de-identify particular fields, that effort is built into the cost.
Recurring cycles
Most partnerships run on a recurring basis rather than a one-time export. A lab that sends us a fresh batch each month or quarter gets paid each cycle. The amount per cycle stays in the same range and adjusts as the data evolves.
Want a rough estimate for your institution? Tell us about your data and we'll come back with a number.