Unit Economics and Contribution Logic

Why unit economics discipline matters

Growth covers many sins until it does not. A company can post impressive revenue charts while each new customer makes the P&L worse. Unit economics is the discipline of asking, for each incremental customer, order, or transaction: does this unit strengthen or weaken the business once variable costs and acquisition costs are counted?

Venture-backed companies sometimes subsidize negative unit economics for years, betting that scale, pricing power, or cost learning will flip the sign later. Operators without that luxury cannot. Even funded companies eventually face the same math when capital markets tighten. Boards that ignored contribution logic during boom years ask painful questions during downturns: "Why did doubling customers double losses?"

From Lesson 2, you separated COGS, OpEx, and capex and built a unit P&L. This lesson goes deeper: contribution margin, CAC (customer acquisition cost), LTV (lifetime value, cumulative contribution over customer life), payback period, and cohort thinking. You will work full numeric examples with checks, because vague "we are roughly healthy" statements fail when churn moves a point or CAC rises 20%.

Contribution margin: the atomic unit of profit logic

Contribution margin equals revenue per unit minus variable costs per unit. Variable costs include direct COGS, payment processing, shipping, variable support when tied to usage, and sales commissions paid on the sale. Contribution margin pays for fixed OpEx and profit. It deliberately excludes allocated headquarters rent debates until you know the unit works.

Why exclude fixed costs at first? Because fixed costs are choices about scale and ambition. A negative-contribution customer cannot be "saved" by clever allocation. A positive-contribution customer may still be worthless if CAC is huge and life is short, but you diagnose in layers.

A direct-to-consumer coffee subscription might show: $28/month revenue, $9 beans and packaging, $0.81 payment processing (2.9%), contribution $18.19/month. That $18.19 is what pays email tools, brand ads (fixed in short run), and founder salary. If CAC was $45 and average life is 8 months, cumulative contribution ≈ 8 × $18.19 = $145.52 before fixed corporate costs. Compare to CAC $45: unit looks healthy if churn stays stable.

Managers watch contribution when pricing changes, COGS inflation hits, or support tickets rise with product complexity. A 5% price cut that does not increase retention can destroy contribution faster than finance models if variable costs are sticky.

CAC, LTV, and payback

CAC is the fully loaded cost to win one new paying customer: ads, sales compensation, marketing tools, content production, allocated SDR (sales development representative, outbound prospecting) time, and often a slice of founder selling hours in early stage. Understated CAC is one of the most common startup pathologies.

LTV in disciplined unit economics means contribution LTV, not revenue LTV. Revenue LTV flatters. A customer who pays $100/month but costs $95 to serve has tiny contribution LTV even if revenue LTV looks impressive.

Payback period is how many months until cumulative contribution equals CAC. Until payback, growth consumes cash. After payback, incremental customers fund growth and fixed costs (if retention holds).

Common SaaS heuristics (vary by segment; not laws):

Capital cost matters. In high-rate environments, long payback is expensive. Churn volatility matters: LTV formulas often assume constant churn; a renewal cliff makes LTV fragile.

Simple LTV approximation (subscription, constant monthly churn c):

Average life in months ≈ 1/c (exponential simplification).

LTV_contribution ≈ (monthly contribution) × (1/c).

Example: contribution $18.19, churn 4%/month → life ≈ 25 months → LTV ≈ $454 before discounting. With CAC $45, LTV/CAC ≈ 10 ✓ (strong if churn stable).

Discounting and expansion revenue refine this; the managerial point is comparison of like-with-like contribution, not revenue vanity.

A more conservative discounted LTV applies a monthly discount rate to future contribution when capital is expensive or churn uncertain. If monthly discount rate is 1% and contribution is $18.19 for 25 months, undiscounted LTV is $454.75; discounted present value is lower but still comparable to CAC if the same discount is applied consistently across channels. The exact formula matters less than the discipline: do not compare optimistic undiscounted LTV from sales to conservative CAC from finance.

Fully loaded CAC should include creative production, agency fees, event sponsorship amortized over expected leads, and sales engineer time on failed deals. Partial CAC loading is how companies discover at Series C that payback is twice what Series A decks claimed. Document CAC assumptions in the same dashboard as churn; when either moves, rerun LTV/CAC before approving spend increases.

Payback period interacts with cash runway. A company with 12 months runway and 18-month payback cannot grow through paid acquisition without new capital, even if LTV/CAC looks attractive on a three-year horizon. Operators must bridge unit economics to the cash flow statement, not only to investor heuristics.

Churn, expansion, and cohort thinking

Logo churn loses customers. Revenue churn loses dollars from downgrades and cancellations. NRR captures expansion minus churn on existing cohort revenue: NRR 130% means the cohort from last year now pays 30% more in aggregate even after losses.

Blended company averages hide channel disasters. Paid social CAC $200 with 6-month life vs organic CAC $20 with 36-month life blended into "CAC $110" misleads. Cohort analysis tracks customers acquired in the same month or channel through time. Unit economics must be cohort-based for decisions.

Annual prepay can hide churn. Customers paid upfront look retained until renewal cliff. Finance sees cash; customer success sees usage collapse. Model renewal risk explicitly.

Cohort tables track customers acquired in the same month through months 1, 2, 3, etc. Healthy SaaS cohorts show revenue retention stabilizing after an early dip. Unhealthy cohorts show continuous decay, meaning new customers merely refill a leaking bucket. Marketplace cohorts track orders per buyer and take rate per order over time, not only "active users" counts that hide one-time promo riders.

Expansion revenue (upsell, cross-sell, seat growth) raises NRR above 100% and can rescue mediocre initial contribution if onboarding lands customers small and grows them. Enterprise SaaS often underwrites negative first-year contribution on the assumption that year-two expansion doubles account value. That bet requires customer success capacity and product depth; otherwise expansion is a spreadsheet assumption, not a cohort outcome.

Segment unit economics when buyer type differs. A self-serve SMB customer with $200 CAC and $50/month contribution is not comparable to an enterprise logo with $40,000 CAC and $8,000/month contribution after six-month implementation. Blending them produces a fictional "average customer" that no channel actually acquires. Lesson 4 buyer power often explains why enterprise segments tolerate longer payback if expansion and switching costs are real.

Break-even at unit and company level

Unit break-even: contribution per unit > 0. Surprisingly often violated in marketplaces subsidizing both sides.

Company break-even: total contribution covers fixed OpEx.

Breakeven units ≈ Fixed OpEx ÷ contribution margin per unit.

Adding sales headcount (fixed) without improving contribution per customer raises breakeven unit count. Hiring must connect to either higher conversion (lower CAC), higher price, lower variable cost, or better retention (longer life).

Sensitivity analysis managers should run routinely

Before approving a marketing budget increase, rerun three scenarios on a spreadsheet: CAC +20%, churn +1 absolute point, contribution −10% (COGS inflation). If LTV/CAC falls below 2× in the combined downside case, growth spend is fragile. This takes minutes and prevents quarters of blame after macro or competitive shifts.

Unit economics also gate product decisions. A feature that raises hosting cost $2 per user per month must either increase willingness to pay, retention, or viral acquisition enough to pay back. Product teams without contribution visibility ship "success" that finance experiences as margin collapse.

RoastBox ships coffee monthly (fictional). We build unit economics end-to-end.

Part A: Per-unit monthly economics

Contribution margin % = 18.19 ÷ 28 = 65.0%.

Part B: CAC and LTV

Fully loaded CAC (blended): $45 per new subscriber.

Monthly logo churn: 4% → average life ≈ 1/0.04 = 25 months (simplified).

LTV_contribution ≈ 25 × $18.19 = $454.75.

LTV/CAC ≈ 454.75 ÷ 45 = 10.1× ✓

Payback months ≈ CAC ÷ monthly contribution = 45 ÷ 18.19 = 2.5 months ✓

Part C: Company breakeven (monthly)

Fixed OpEx: marketing overhead (non-performance), HQ, tech = $95,000/month.

Contribution per subscriber $18.19.

Breakeven active subscribers ≈ 95,000 ÷ 18.19 ≈ 5,223 subscribers (ignoring new CAC cash timing).

If active base is 6,000, operating contribution ≈ 6,000 × 18.19 = $109,140; after fixed OpEx ≈ $14,140/month operating cushion before new customer CAC investment.

Part D: Managerial read

Investor question: "Can you 3× subscribers?" Operator answer requires cash model: each new sub pays back CAC in ~2.5 months but consumes $45 upfront. Growth of 1,000 net new subs requires ~$45k CAC cash plus variable COGS scaling smoothly. If churn rises to 6%, life ≈ 16.7 months, LTV ≈ $304, LTV/CAC ≈ 6.8× still ok but deteriorating ✓ sensitivity matters.

QuickBite matches diners and restaurants (fictional). Unit = one delivered order.

Part A: Per-order economics

Part B: New diner CAC and payback

Promo spend to acquire new diner: $30 (first-order discounts + ads).

Orders per diner per month (avg): 2.2.

Monthly contribution per diner ≈ 2.2 × $3.30 = $7.26.

Payback on promo CAC ≈ 30 ÷ 7.26 = 4.1 months on variable contribution alone.

If diner retention beyond month 1 is weak (30% one-and-done), effective payback worsens because later orders never arrive.

Check: 2.2 orders × $3.30 = $7.26 ✓

Part C: Subsidy stress

If competitor war raises diner promo to $45 and take rate cannot rise without restaurant revolt, payback ≈ 6.2 months. With 6-month average active life, LTV_contribution ≈ 6 × $7.26 = $43.56 vs CAC $45 → LTV/CAC ≈ 0.97× (value destructive before fixed ops).

Board conversation shifts from "growth rate" to model redesign: membership fee, reduce subsidy, increase order frequency through workplace catering, or exit low-density cities.

Part D: Operator takeaway

Marketplace unit economics must be computed per side and per order. Growth that buys unprofitable orders is deferred pain. Lesson 1 canvas blocks revenue streams (take rate + fees) and cost structure (driver incentives) must move together.

Part E: Frequency lever (second-order unit economics)

If QuickBite raises order frequency from 2.2 to 3.0 orders/month per active diner through workplace lunch bundles (same contribution per order $3.30), monthly contribution per diner becomes $9.90. On CAC $30, payback falls to 3.0 months without changing take rate. This illustrates that marketplace fixes are not only subsidy and take rate; job redesign (capture more occasions) moves unit economics. Operators should model frequency and retention before declaring a city "won."

Connecting unit economics to the canvas (Lesson 1 integration)

Each canvas block influences unit math. Value proposition strength affects churn and willingness to pay (revenue per unit). Channels determine CAC and sometimes variable support cost (enterprise vs self-serve). Revenue streams determine whether expansion shows up in NRR or requires new sales. Key activities like white-glove onboarding add semi-variable labor that beginners forget when they only subtract hosting COGS.

When presenting unit economics to executives, tie numbers to canvas blocks so leaders know which block to fix. A low LTV/CAC from high churn is a delivery and value proposition problem, not a marketing efficiency problem. A low LTV/CAC from high CAC is channel and sales model problem. Misdiagnosis wastes quarters.

Investors sometimes ask for contribution margin after CAC in month zero (immediate unit profit including acquisition). Rare in high-growth B2B, but useful in transactional models where repeat purchase is same-day or same-week. Food delivery and convenience marketplaces sometimes achieve fast repeat, making first-order contribution after promo subs critical. Always match the unit time horizon to the customer behavior of the category.

For board reporting, show cohort LTV curves alongside blended LTV. Blended averages age poorly when mix shifts from enterprise to SMB or from organic to paid. A rising blended LTV can mask worsening paid cohorts if organic cohorts improve simultaneously. Separation keeps capital allocation honest.

Finally, document the unit explicitly in every model: per customer, per order, per seat-month, per trip, per gigabyte-month. Teams that debate LTV without agreeing on the unit talk past each other. The unit should match how the customer buys and how COGS scales, linking Lessons 2 and 3 in one cell on the spreadsheet.

When churn is quoted, specify logo churn vs revenue churn vs net revenue retention. A stable logo count with shrinking revenue per logo signals downgrades or reduced usage, not health. Unit economics without revenue churn misses the downgrade path that enterprise SaaS faces when customers consolidate vendors during downturns. Build both views into the same monthly operating review so finance and customer success share one definition of health.

StudyFlow sells B2B study tools to university departments. Per department contract: $1,200/month; hosting and support variable $180/month; implementation one-time fee $2,400 with $1,100 contractor cost (only on new logos). Sales cycle 3 months; AE cost allocated $9,000 per closed logo. Average department life 30 months before churn; monthly churn constant approximation 3.3%.

1. Compute monthly contribution margin per active department.
2. Compute CAC per new logo (include implementation cost and AE).
3. Estimate LTV_contribution and LTV/CAC.
4. Payback period in months from CAC using monthly contribution only (ignore implementation fee timing).
5. Explain one lever that improves payback without cutting price.

Solution

1. Monthly contribution:

$1,200 − $180 = $1,020/month.

2. CAC:

Implementation variable $1,100 + AE $9,000 = $10,100 (implementation is cost to win/onboard new logo).

Some firms treat implementation as COGS of sale; here included in acquisition economics.

3. LTV:

Life ≈ 1/0.033 ≈ 30.3 months (consistent with stated 30).

LTV_contribution ≈ 30 × $1,020 = $30,600.

LTV/CAC ≈ 30,600 ÷ 10,100 = 3.03× ✓ (workable if churn stable).

4. Payback:

10,100 ÷ 1,020 ≈ 9.9 months on subscription contribution alone.

One-time implementation fee $2,400 with $1,100 cost adds $1,300 contribution at sale → adjusted first-month boost reduces payback slightly if cash collected upfront:

Effective CAC net of implementation contribution = 10,100 − 1,300 = $8,800.

Payback ≈ 8,800 ÷ 1,020 ≈ 8.6 months ✓

5. Lever (prose):

Improve retention (reduce churn) lengthens life and LTV faster than price cuts. Alternatively, reduce AE cost via inside sales for smaller departments, or productize implementation to cut $1,100 contractor COGS. Raising price $100/month adds $100 contribution and cuts payback ~0.8 months if churn unchanged.

GlowBox beauty subscription: $35/month; COGS $12; processing $1.12; CAC $52; churn 5%/month.

1. Compute contribution, LTV_contribution (use 1/churn life), LTV/CAC, payback months.
2. GlowBox tests a $5/month price increase; churn rises to 5.5%. New contribution? New LTV/CAC? Should they keep the increase? Explain in a paragraph.

Solution

1. Baseline:

Contribution = 35 − 12 − 1.12 = $21.88.

Life = 1/0.05 = 20 months.

LTV = 20 × 21.88 = $437.60.

LTV/CAC = 437.60 ÷ 52 = 8.42×.

Payback = 52 ÷ 21.88 = 2.38 months ✓

2. Price increase scenario:

New price $40; assume processing ≈ $1.27 (2.9%).

Contribution = 40 − 12 − 1.27 = $26.73.

Life = 1/0.055 ≈ 18.18 months.

LTV = 18.18 × 26.73 ≈ $486.0.

LTV/CAC ≈ 486 ÷ 52 = 9.35× ✓ Still improves despite churn rise.

Payback = 52 ÷ 26.73 ≈ 1.95 months.

Keep increase if churn stabilizes at 5.5% and brand damage is contained. Higher LTV/CAC shows price elasticity favorable; monitor cohort retention qualitatively for complaint spikes.

Key takeaways

• Contribution margin is the first test; LTV/CAC and payback come next.
• Always use contribution LTV, not revenue LTV.
• Cohort and channel specificity beats company averages.
• Marketplace and subscription math must include both sides and renewal risk.
• Growth without positive unit economics is deferred pain.

1. Build a one-page unit model for an offer you know: contribution, CAC, LTV, payback.
2. Identify where your organization most likely understates CAC or overstates LTV.
3. Continue to Lesson 4: Business Model Fit with Industry Structure.