all

wiki/concepts/unit-economics.md

@@ -0,0 +1,59 @@
+# Unit Economics
+
+#concept #pricing #finance
+
+## Summary
+
+The full-cost arithmetic beneath a price — what it actually costs to win, serve, and *keep* a customer, and what margin must remain. Founded by [[2026-07-26-how-to-build-a-billion-dollar-company-2027]] ([[oskar-hartmann]]), this page **partially fills a gap the vault's own meta-analysis flagged from the start** ("B2B unit economics — CAC/LTV — unfilled by the corpus"). Partial, because the source's principles are general and its examples consumer-grade; the B2B-services version is still missing. `Status: tentative` — single author (two sources since 2026-07-26: the Zavent post-pay trap and the CAC-burn restatement are the same voice, so consistency rather than corroboration).
+
+## Current Understanding
+
+**The central diagnosis: most entrepreneurs sell below the real cost.** Not below the *visible* cost — below the full one. Systematically forgotten lines:
+
+- distribution and sales cost (the channel itself is a cost of goods — [[sales-channel-as-moat]])
+- **repeat** acquisition and retention (winning the customer once is not winning them)
+- transport, amortization
+- inventory write-offs
+
+The resulting "cash gaps" founders explain away as timing are, in Hartmann's telling, **real losses** that were priced in from the day the price was set. This is [[pricing-from-value]]'s "sell dear" rule arrived at from the cost side rather than the value side — an independent tradition converging on the same instruction.
+
+**The sequence:** (1) prove the product is *needed*; (2) prove you can produce it **far below** willingness-to-pay; (3) the spread must cover *everything* — marketing, distribution, sales, warehouse — **with a buffer**. Note the order: demand first, cost structure second, price last. Compare [[productized-service]]'s validate-before-build rule — same discipline, one level deeper.
+
+**The buffer rule / best-year fallacy:** never take your best year as the base. The best year is a once-a-decade anomaly; budget from it and every normal year reads as a crisis. Outside razor-thin retail, margin must carry a buffer.
+
+**LTV → CAC → moat:** a business that has run its channel long enough to *know* lifetime value can pay up to **~⅓ of lifetime profit** for a customer on day one (US credit-card customer ≈ **$1,000 CAC**; 1,000 customers = $1M — "it doesn't come cheaper"). That spending level is an **entry barrier**: competitors without the LTV statistics can't rationally match the bid. This is also the honest explanation of loss-making venture rounds — they fund negative unit economics until the LTV arrives ([[venture-fit]]).
+
+**Post-payment turns you into your customer's bank** ([[2026-07-26-main-principle-of-successful-business]], added 2026-07-26 — same author, second source). The Zavent anti-case: demand looked strong, but customers would only pay **3 months after delivery**, so the company was silently financing its clients — a working-capital cost that belongs on the forgotten-lines list above. When prepayment was required, conversion collapsed, revealing the real demand level. Twin lessons: **prepayment willingness is itself a unit-economics variable** (it decides whose balance sheet carries the build — [[sell-before-build]]), and payment *timing* is part of the full cost of a sale.
+
+**The CAC-burn barrier, restated** (same source): once PMF is proven and unit economics are positive, deliberately spend heavily per customer ($80–100 CAC in his example) so no newcomer can afford to enter. Consistent with the ⅓-of-LTV rule from his first source — framework stability for this author, not corroboration.
+
+**Pricing power as the PMF test** (recorded in full on [[pricing-from-value]]): if raising prices doesn't shrink the customer flow, you have PMF; if a marketplace sets your price, "you're in a simulation of entrepreneurship." Unit economics you don't control aren't yours.
+
+## Evidence
+
+- Full-cost list, cash-gaps-are-losses, the three-step sequence, buffer/best-year rule, LTV×⅓ CAC, $1,000 credit-card CAC, entry-barrier logic — [[2026-07-26-how-to-build-a-billion-dollar-company-2027]]
+- Zavent post-payment trap ("a bank financing its clients"); $80–100 CAC-burn barrier restated — [[2026-07-26-main-principle-of-successful-business]] (same author — consistency, not corroboration)
+- Convergent from the value side: price must fund the work that makes the service good ("below ~$100/mo there's no margin…") — [[2026-07-17-design-the-perfect-offer]], [[pricing-from-value]]
+- Convergent: Solution-model margin 30–50% vs staff-aug rate race — [[solution-vs-staff-augmentation]]
+- The gap this partially fills — flagged in [[2026-06-15-meta-analysis-selling-dev-in-ai-era]]
+
+## Related Pages
+
+- [[pricing-from-value]] — the value side of the same price; pricing power lives there
+- [[sales-channel-as-moat]] — the channel whose cost and payback this math governs
+- [[venture-fit]] — negative unit economics as a deliberate, funded phase
+- [[productized-service]] — validate-before-build is step 1 of the sequence here
+- [[marketing-system]] — "money in → more money out" is a unit-economics statement
+- [[overview]]
+
+## Contradictions / Uncertainty
+
+- `Status: tentative` — one source; every number ($1,000 CAC, ⅓-of-LTV) is a stage figure without citation.
+- **The examples are consumer/product; the vault's domain is B2B services.** Services LTV is lumpy (projects, retainers), churn behaves differently, and "repeat acquisition cost" may be the [[referrals]]/[[partnerships]] machinery rather than ad spend. The transfer is plausible but unshown — the flagged B2B gap is *narrowed*, not closed.
+- **Tension with the vault's spend-nothing school:** Martell's $0 blueprint says spend nothing until customers pay ([[sales-discipline]]); Hartmann describes rationally spending $1,000 to acquire one customer. Reconcilable as stages (pre-LTV-knowledge vs post-), but that seam is exactly what neither source specifies.
+
+## Next Questions
+
+- What are CAC and LTV for a niched dev-services operator, concretely — and does the ⅓ rule mean anything when LTV is 2–3 projects?
+- Where is the line between a venture-fundable negative-unit-economics phase and Fab.com-style self-deception ([[venture-fit]])?
+- Which of the forgotten cost lines apply to services (repeat acquisition, surely; write-offs — as unbilled rework?)?