"Is it Cake? (Revenue Edition)" — CJ Gustafson (Mostly Metrics)
Why this is in the vault
The piece is both a useful operating reference (six-level revenue hierarchy I can plug into Squarely / MAC / Sanity Check unit economics without re-deriving) and a primary-source voice study for Sanity Check. CJ's "is it cake" pop-culture handle for "real recurring revenue vs. plausible-looking imposters" is exactly the cadence I've been told the founder wants from SC: a familiar reference repurposed to bite. Filing for both functions.
⚠️ Sponsorship
Sponsored by Abacum AI (FP&A automation). Summit recap at top + interactive demo link. Same advertiser-of-record as the May 10 token piece — Abacum is running a multi-issue sponsorship. Doesn't bias the revenue-hierarchy content (which is operator-evergreen), but worth tracking that CJ's recent cadence is heavily Abacum-anchored.
Issue contents
Thesis essay + numbered framework + bonus data charts (Mostly Multiples). Hybrid format — practitioner essay with embedded benchmarking sidebar.
Core thesis
Early-stage founders inflate revenue not from malice but from accounting illiteracy. CJ lays out a six-level hierarchy from "indications of interest" through "true recurring revenue" so operators (and the people evaluating them) can pinpoint exactly which level a number lives at. The fix is taxonomy, not enforcement.
The six-level revenue hierarchy
| Level | What it is | Why it isn't real revenue |
|---|---|---|
| 0 | Indications of interest | Verbal, non-binding |
| 1 | Non-binding LOIs | Paper exists, no obligation |
| 2 | Negative/positive GMV | Marketplace flow, not yours |
| 3 | Bookings | Contracted but not yet earned |
| 4 | Transactional revenue | One-time, not recurring |
| 5 | Recurring (subscription) | Recurring but not necessarily renewing |
| 6 | True recurring revenue | Renewing cohort with retention proof |
Other frameworks
- cARR ≠ ARR. Multi-year ramped deals inflate cARR by 40%+ over actual cash collection in year one. The gap is where founders unintentionally mislead.
- Operating leverage target: +25% profitability through declining OPEX across four buckets (COGS, S&M, R&D, G&A).
- Rule of 40 + RPE benchmark ($595k revenue per employee as a current public-tech median).
- 24-month CAC payback as the current SaaS median (worse than the 18-month "healthy" target many decks still cite).
Voice tactics worth stealing (for Sanity Check)
CJ has two moves I want to study and lift:
- Self-deprecating-while-authoritative. The shark-shirt anecdote ("getting my ass handed to me, all I could think about was how dumb this shark shirt must look") establishes that he's been the wrong-side-of-the-table guy. Then he pivots to the framework. This is the inverse of LinkedIn "humble brag" — it's "actual deprecation, then expert verdict." Sanity Check should run more of this. The founder already does it naturally in his X posts.
- Permission-to-disagree. "Which brings me to something most finance people won't want to hear" telegraphs that the next sentence is contrarian. It primes the reader to brace for a take, which is the opposite of newsletter pap. Pattern: signal the disagreement before delivering it.
Mapping against Ray Data Co
- Direct framework lift for MAC. When MAC pitches to revenue-stage founders, "where on the six-level hierarchy is your number" is a better discovery question than "what's your ARR." I should re-write the MAC discovery script to use this taxonomy.
- For Squarely specifically. Squarely's revenue is currently Level 4 (transactional, App Store purchase). The honest framing for any pitch deck is L4-with-attach-rate-experiment-to-L5. Treating it as L5/L6 would be the inflation pattern CJ is calling out.
- Sanity Check piece candidate. "Six Levels of Revenue and Why Founders Pick Level 6 by Default" — angle is the social-pressure reason inflation happens, not the accounting reason. Different enough from CJ's framing to not be derivative. Logging as candidate, not as commitment.
- The shark-shirt move is teachable. Voice-match skill should have a specific "self-deprecating frame to expert verdict pivot" sub-pattern. Adding to my notes for the next voice-study pass.
Related
- [[2026-05-10-mostlymetrics-token-budget-as-employee-cost]]
- [[../01-projects/sanity-check/voice-study]]
- [[../01-projects/mac/README]]