Two numbers that appear on nearly every board deck
The first is an LTV to CAC ratio of at least 3:1, presented as the threshold for healthy unit economics.
The second is that a 1% improvement in price produces roughly an 11% increase in operating profit, presented as evidence that pricing is the most powerful profit lever available.
Both numbers have real origins. Neither means what it is usually taken to mean, and a business model designed around either without understanding its conditions will produce confident decisions on unstable ground.
This article covers what a business model has to establish, and how to handle the two metrics that carry most of the weight in that conversation.
What a business model has to answer
A business model describes how value is created, delivered, and captured. Three verbs, and the third is where most early-stage work is thin.
Six questions, and a model is incomplete until all six have answers that agree with each other.
Who pays. Not who uses. In many models the user and the payer are different, and the model breaks if that distinction is not explicit.
For what unit. Per seat, per transaction, per outcome, per period, per volume. The unit determines how revenue scales relative to cost, and changing it later is a repositioning exercise, not a pricing tweak.
On what basis is the price set. Cost-plus, competitor-referenced, or value-based. Most businesses claim value-based pricing and practise competitor-referenced pricing.
What does it cost to serve. Fully loaded, including support, fulfilment, payment processing, returns, and the fraction of the team the customer consumes.
What does it cost to acquire. Fully loaded sales and marketing for a period, divided by the customers that spend produced. Not just media cost.
How does the customer behave over time. Repeat rate, churn, expansion. This is the input everything else depends on, and the one most early-stage businesses have the least data for.
The Business Model Canvas is a reasonable way to lay these out in one place. It is an artefact for surfacing disagreement, not an analysis. Nine boxes of assertions do not test whether the model works.
The 3:1 rule, and where it came from
The LTV to CAC ratio of 3:1 traces to David Skok at Matrix Partners, published around 2010 in SaaS Metrics 2.0 on the For Entrepreneurs blog.
His framing was explicitly a guideline: that for a successful SaaS business, this number should be higher than three. He also noted that the strongest SaaS businesses he had observed ran considerably above it, sometimes seven or eight times.
The conditions underneath matter more than the number. Skok derived it from observing mature public SaaS companies — HubSpot, Salesforce, NetSuite — at steady state, with stable churn, multi-year customer lifetimes, and CAC payback comfortably inside twelve months.
Fifteen years later it appears in seed decks written by companies meeting none of those conditions.
OpenView's Blake Bartlett has put the objection directly: nobody knows why three is the appropriate benchmark. It simply is. His specific criticism is that the standard formula assumes constant churn and a finite customer lifetime, which breaks down for genuinely retention-strong businesses.
Under the evidence grading used across this series, 3:1 is a documented practitioner guideline that has been repeated as though it were derived from a broad dataset. It is a starting point, not a law.
Three ways the ratio misleads
The LTV input is unstable. Different calculation methods produce materially different answers from the same data — the same customer can come out at very different values depending on which method is used. Failing to margin-adjust overstates LTV substantially, and an overstated LTV authorises acquisition spend the business cannot actually afford.
The SaaS formula is misapplied to non-SaaS models. SaaS LTV is gross-margin-adjusted recurring revenue across a multi-year subscription. Direct-to-consumer economics have a fundamentally different cash profile — a one-to-two purchase repeat curve rather than a multi-year contract. For DTC, a healthy range sits nearer 2.5:1 to 4:1 measured on a twelve-month verified cohort basis, with lifetime value expressed as contribution margin rather than revenue.
A high ratio can be a warning. Above roughly 5:1, the more likely reading is under-investment in acquisition than exceptional efficiency. The business could profitably spend more and is not.
The more defensible alternative: CAC payback period. It depends only on acquisition cost and current gross profit per customer, with no churn assumption and no lifetime projection. For a business without stable cohort data — which describes most early-stage companies — payback is the number to run the model on.
The pricing claim, handled properly
The 11% figure comes from Michael Marn and Robert Rosiello's 1992 Harvard Business Review work at McKinsey. A later McKinsey formulation, using the average S&P 500 income statement, puts the same effect nearer 8%.
Here is the part that is almost never stated: it is not a market experiment. It is an arithmetic exercise. Take current margins, raise price by 1%, hold volume constant, and observe the effect on operating profit. Because price flows entirely to the bottom line while costs are a fraction of revenue, the result is arithmetically guaranteed to be large.
The assumption doing all the work is that volume stays constant. In real markets, raising price reduces unit sales. The size of that reduction is the entire question, and the calculation does not address it.
So the honest version: price is a high-leverage input, and the 11% figure demonstrates the leverage rather than measuring the outcome. Simon-Kucher's 2025 global pricing study found volume's influence on profitability declining — from around 50% in 2021 to 40% in 2025 — with pricing remaining powerful and under-used.
What follows practically is not "raise prices." It is that pricing deserves the same testing discipline as any other high-leverage decision, and that most businesses set price once and revisit it only under pressure.
Where business model design actually fails
Price is set by looking at competitors. This inherits a competitor's cost structure, positioning, and margin assumptions, none of which are yours. It is also the most common approach.
Cost to serve is understated. Support time, returns, payment fees, and fulfilment are frequently excluded from early models, which makes contribution margin look healthier than it is and makes acquisition spend look more affordable than it is.
The model assumes retention it has not observed. Lifetime value calculated from three months of data on a twelve-month projection is a hypothesis wearing a number's clothing.
The unit is wrong for how value accrues. Charging per seat when value scales with transactions, or per transaction when the customer's benefit is availability, creates permanent friction between what the customer perceives and what they pay.
Structural limits are ignored. If the category has never produced the margins the model assumes, the model needs an explicit account of what it will do differently. Structural analysis belongs upstream of the financial model, not after it.
Diagnostic: is the model tested or asserted?
Seven tests.
The payer is named separately from the user.
Cost to serve includes support, fulfilment, payment processing, and returns.
CAC is fully loaded, including salaries and agency costs, not just media spend.
Lifetime value is margin-adjusted, and the calculation method is written down.
CAC payback period is tracked alongside any ratio, and the model works on payback alone.
Retention in the model is observed, not assumed — with the observation window stated.
Price has been tested at least once, rather than set once.
Failing test four or five means the unit economics are directional at best. Failing test seven is nearly universal and is the cheapest thing on this list to fix.
What this produces
A model that states what the business must be true about the world in order to work — and how each of those things would be checked.
That is the practical output of business model design at the Define stage. Not a financial projection, which will be wrong, but a set of named assumptions with measurement attached: this is what a customer costs to acquire, this is what they are worth, this is how long the money takes to come back, and this is the evidence behind each figure.
Those assumptions are what the next phase's budget is built on. A model that has not distinguished between what it has observed and what it has assumed will produce a budget that cannot tell the difference either.
Frequently asked questions
Is a 3:1 LTV to CAC ratio the right target?
It is a guideline published by David Skok around 2010, drawn from mature public SaaS companies at steady state with stable churn and payback inside twelve months. Most companies citing it do not meet those conditions. Treat it as a starting reference and vary it by business model and stage.
What is a healthy LTV to CAC ratio for a DTC business?
Reported healthy ranges for DTC ecommerce sit nearer 2.5:1 to 4:1, measured on a twelve-month verified cohort basis with lifetime value expressed as contribution margin rather than revenue. The SaaS 3:1 rule assumes a multi-year subscription cash profile that DTC does not have.
Why is CAC payback period better than LTV:CAC for early-stage businesses?
Payback requires only acquisition cost and current gross profit per customer. It needs no churn assumption and no lifetime projection, which is exactly what an early-stage business lacks reliable data for.
Does a 1% price increase really produce an 11% profit increase?
The figure comes from a 1992 McKinsey analysis and is an arithmetic exercise, not a market experiment. It holds volume constant, which real price increases do not. It demonstrates that price is high-leverage; it does not predict the outcome of raising yours.
Can an LTV to CAC ratio be too high?
Yes. Above roughly 5:1, the more likely explanation is under-investment in acquisition than exceptional efficiency. The business could probably spend more profitably and is choosing not to.
Is the Business Model Canvas worth using?
As an artefact for getting the whole model visible in one place and surfacing disagreement between leaders, yes. As analysis, no — filling nine boxes with assertions tests nothing.
When should the business model be revisited?
When observed retention differs from the assumption, when cost to serve moves, when the category's structural economics shift, or when a pricing test produces a result. Business models age against reality, not against the calendar.
Sources
Skok, D., SaaS Metrics 2.0, For Entrepreneurs / Matrix Partners (circa 2010) — origin of the 3:1 LTV:CAC guideline
Bartlett, B., OpenView — published criticism of the 3:1 benchmark's derivation
Marn, M. and Rosiello, R., Managing Price, Gaining Profit, Harvard Business Review (1992); McKinsey, The Power of Pricing
Simon-Kucher, Global Pricing Study 2025
Published 2026 LTV:CAC benchmark ranges by business model, including DTC ecommerce and DTC subscription
Osterwalder, A., Business Model Canvas
Structure your next phase
Zerologic works with leadership teams on revenue architecture — pricing logic, unit economics, and the assumptions a growth budget will be built on.
Talk to us: partners@zerologic.io · zerologic.io
