Costs Behaving Badly: Unit Economics I
Why this is in the vault
Part 1 of the Secret CFO's Unit Economics trio (Jul-Aug 2024). Opens with a cold-open disaster: a factory mis-priced a major contract because the costing model was wrong, the contract was loss-making at scale, factory closed, 900 jobs lost. The thesis is that all costs are mixed costs if you look hard enough - the textbook fixed-vs-variable distinction is wrong in practice, costs step at volume thresholds, and contribution margin (behavior-based) is fundamentally different from gross profit (accounting-classification-based). The three-level finance maturity model (accurate cost capture → analytical capability → embedded organizational discipline) is a clean lens for asking "where is RDCO actually breaking down" on any given bet's economics. Most usefully: "fixed costs are 10x easier to increase than to reduce" - a maxim that should govern every RDCO infra commit.
Mapping against Ray Data Co
This is the foundational Unit Economics piece for MAC's ICP - data engineers and finance-tech buyers - and the cold-open disaster is exactly the failure mode MAC sells against. Specific RDCO maps:
- MAC content alignment: the three-level finance maturity model gives MAC a vocabulary for staging the buyer journey. Level 1 (cost capture) maps to data quality at the warehouse layer; Level 2 (analytical capability) maps to dbt-style transformation discipline; Level 3 (organizational embedding) maps to the data-contract / observability harness MAC actually advocates. The article inadvertently writes MAC's content arc.
- Squarely cost behavior: Apple TF render minutes are textbook step costs - flat until I cross a build-volume threshold, then they jump. The "all costs are mixed costs" frame should govern how I budget Squarely's monthly burn (use cell ranges, not point estimates).
- RDCO infra commits: the "10x easier to increase than reduce" maxim is the rule I should apply to any new SaaS / agent-tier commit. Once Anthropic spend, ElevenLabs commit, or any monthly infra is on, getting it off requires founder-level intervention.
- Sanity Check sponsor pricing (eventual): when SC ever takes paid sponsorship, this is the article that explains why the right price is contribution-margin-driven, not gross-margin-driven.
⚠️ Sponsorship
Sponsored by Orb (usage-based billing platform). Topic-matched - usage-based billing IS the operationalization of variable-cost-revenue alignment, which is the article's thesis. Disclosure clean. No editorial bend toward Orb's product specifically.
Related
- [[06-reference/2024-07-13-cfosecrets-7-levers-pl-unit-economics-ii]] - Unit Economics II, the 7 P&L levers companion
- [[06-reference/2024-08-17-cfosecrets-strategic-unit-economics-iii]] - Strategic Unit Economics, the strategic-finance-framed trio capstone
- [[06-reference/2026-04-30-mac-bet-architecture-audit]] - MAC bet architecture; this article writes MAC's three-level content arc
- [[06-reference/2026-05-02-cfosecrets-working-capital-warfare-i-cash-conversion-cycle]] - Working Capital Warfare I; the operating-cycle companion