We sit in data rooms where the spreadsheet looks beautiful and the assumptions collapse in minute four of live Q&A. Before every Series A we run, we insist the founder can defend three artefacts in under ten minutes each—with numbers tied to product reality.
Model one: unit economics the way a skeptical partner asks
Not LTV/CAC as motivational poster. CAC by channel, payback months with gross margin after deliverables, churn defined the way finance and product both sign. We print the bridge from last month’s cohort to this month’s.
If the answer to “what breaks this” is silence, the model is decorative.
Model two: cash, not runway theatre
Bank cash → collections → payroll and infra with dated hiring rows. We expose where the model assumes price increases, payment term changes, or prepaid pull-forwards. Investors smell those fast; we name them first.
Model three: two scenarios and a stress
Base reflects the pipeline the sales lead actually believes. Upside ties to one distribution bet. Stress picks an ugly pair: delayed enterprise close + higher CAC. Not twenty macros—three pictures the board can repeat.
| Question we rehearse | Pass / fail signal |
|---|---|
| “Show me revenue this month from existing customers only.” | Drill takes < 30 seconds |
| “Where is churn concentrated?” | Segment known |
| “What hiring row goes if Q3 slips two weeks?” | Named role |
Where this breaks down
When product changes price and nobody updates the unit sheet for six weeks. When “ARR” includes pilots with no written renewal path.
Our read
The models are not magic. The rehearsal is. If you cannot explain the bridge out loud, the Excel is not ready—no matter how pretty the chart.