The Situation
The startup’s prior cost model had been built on aspirational unit economics that wouldn’t survive a serious investor question. Manufacturing salaries were out of line with what cell stack assembly actually required. The split between in-house manufacturing and outsourced fabrication was muddled. Tax and benefits assumptions on salaries didn’t match the actual payroll registers.
Investors were going to pressure-test the cost line. The team needed COGS that would hold up.
What CEI did
Recalculated manufacturing salaries based on stacks-per-unit, conservative assembly time estimates, generous hourly rates, and explicit early-stage inefficiency allowance.
Separated the in-house cell stack assembly economics from the outsourced fabrication economics — with a documented markup on the outsourced spend rather than an opaque blend.
Built the salary tax and benefits layer realistically: matched 2025 actuals from payroll registers, ramped to 25% over a multi-year horizon as benefits coverage was added, with the rationale documented.
Created the COGS by component, by year, with sensitivity bands on each material cost decline curve so investors could pressure-test each assumption separately rather than the model as a whole.
Component-Level
Cost Build Documented
Pressure-Tested
Sensitivity Bands per Driver
Defensible
Investor-Ready COGS Model
Investors got a cost model they could actually engage with — not a black box to argue with. The technical and finance teams aligned on a single set of operating assumptions instead of carrying two competing views into investor meetings.
In Their Words
“Our old cost model was a story. CEI built one with math. That is what diligence actually requires.”
— Founder | Confidential Client