An honest comparison of the StromaOS operating model against the bundled data-plus-services stack. Your data license is held constant on both sides - because you bring your own - so this compares the layer where the difference actually lives: the labor and the tooling.
Running an evidence operation costs you three things: a data license, the labor that turns data into evidence, and the tools that govern and stitch it together. StromaOS doesn't touch the first - you keep your license and pay it either way. It replaces the second and third: elastic governed execution instead of a fixed services markup, and one governed platform instead of a stack of point tools.
Start from how you run evidence operations today, then set your program. Defaults describe a mid-size biopharma running ~20 studies a year across two production domains. Every figure is editable; the panel recalculates as you type.
The domains where governed work executes, the Intelligence Modules you subscribe, and how you deploy. Runtime is domain-blended - you pay for the domains that do real work, not for agents, seats, or personas.
Intelligence Modules are a flat subscription; runtime is metered on the domains where governed work runs, scaled by intensity and by the months it's active, with standby at a 10% floor. Building agents, fleets, and swarms is free - you're billed only for governed execution. Air-gap applies a deployment multiplier and a one-time first-year setup; in air-gapped deployment, standby bills as active because it can't be verified without telemetry.
The labor and tooling layer you carry now, on top of your data license. These are the lines StromaOS is designed to replace - edit them to match your actuals or a quote you hold.
StromaOS is not a data business and doesn't resell or replace your entitlements. Bringing your own license means the data cost is identical whether or not you adopt StromaOS - so putting it in the comparison would only pad the number on both sides and net to zero. In a typical bundled contract the data license is 40–50% of spend, normalization 15–25%, and the analytics platform plus expert labor 30–40%: StromaOS leaves the first untouched and targets the last two - elastic governed execution instead of a fixed 1.5× services markup, and one governed plane instead of a stack of point tools and the integration stitched around them. If your inputs don't match your environment, replace them - the defaults are deliberately conservative.
A visible formula is not the same as a sourced one. Each default below is labeled with its basis so you can decide which to keep and which to replace with your own figures before you take the output anywhere.
Every value in this table is editable in the calculator above. None is a quote, and none is required for the model to run.
| Assumption | Default | Basis | Classification |
|---|---|---|---|
| Analyst and services spendAnnual, per production domain | $950,000 | Composite of analyst headcount plus contracted overflow for a mid-size program running roughly 20 studies a year.Replace with your own contracted spend - this is the single largest driver of the result. | Internal estimate |
| Data license shareOf total bundled spend | 40–50% | Held identical on both sides of the comparison and excluded from the delta, because you bring your own license either way.Its accuracy does not affect the saving shown. | Internal estimate |
| Normalization share | 15–25% | The portion of bundled spend attributable to ingestion, mapping and re-aggregation. | Internal estimate |
| Analytics and expert labor share | 30–40% | The portion attributable to the analytics platform plus the expert labor operating it. This is the layer StromaOS is designed to replace. | Internal estimate |
| Services markup | 1.5× | Applied to analyst and overflow labor to reflect contracted rather than internal delivery.Fixed in the model; edit the underlying spend to reflect a different markup. | Illustrative |
| Standby floor | 10% | Minimum billing for a domain held available but not executing. | Product behavior |
| Program size | ~20 studies/yr | Sizes a mid-size biopharma running two production domains. Heavier or lighter programs move every figure. | Illustrative |
If the defaults don't match your program, or you're putting this in front of a CFO, we'll build a tailored model against your real domains, study calendar, and any quotes you hold. We'd rather have the specific conversation than hand you a generic number.