Startup unit economics: metrics, formulas and a worked example

Unit economics shows whether a business makes money on a single customer or sale.

Its core is the customer acquisition cost (CAC), the margin per customer and the lifetime value (LTV).

A model is healthy when LTV is well above CAC and acquisition costs pay back quickly. If the unit economics do not work, growing sales only speeds up losses.

Why do many startups fail even with a strong idea and energetic founders?

Inspiration and enthusiasm are good, but without a precise understanding of the project’s economics, all the effort can end where it began — on a beautiful pitch deck.

Unit economics is not a buzzword from startup manuals; it is the foundation a business stands or falls on.

Below: what it is, which metrics to calculate, a full worked example, common mistakes and what to show investors.

What is unit economics

It is a way to understand one simple thing: does the startup make money on each customer, or lose money even while total revenue grows?

Unit economics answers what happens to the project’s finances at the level of one sale, one customer, one subscription.

It is not bookkeeping that records facts; it is a model of reality that shows whether the business idea is viable at all.

If the unit economics do not work, scaling only increases losses — like pouring water into a leaky barrel: the more you pour, the faster you lose it.

The metrics of unit economics

An important note on LTV. If LTV is calculated on revenue, the model looks better than it is: the customer’s revenue still has to pay for cost of goods, fees and delivery.

It is more accurate to compare margin with CAC: LTV = margin per customer per period × average number of periods, and with constant churn the average number of periods ≈ 1 ÷ churn per period.

Revenue is taken excluding VAT: the tax belongs to the tax office, not the business.

A worked example

A subscription service for small businesses (illustrative figures):

• subscription — €49 a month excluding VAT;
• variable costs per customer — €12 a month: hosting, payment fees, support;
• CAC — €180;
• churn — 4% of customers a month;
• fixed costs — €18,000 a month.

A common investor rule of thumb is LTV at least three times CAC and acquisition payback within a year.

These are guidelines, not laws, but the model in the example clears them comfortably.

With 10% monthly churn, however, a customer would stay 10 months, LTV would fall to €370 and LTV/CAC to 2 — one parameter turns the whole picture around.

Enter your own numbers — the calculator works out margin, LTV, LTV/CAC, CAC payback and how many customers cover your fixed costs. The starting values are the example above.

CAC: formula and benchmarks

CAC (customer acquisition cost) is the average cost of winning one new paying customer.

Investors check it against LTV and margin: it shows whether growth pays for itself or every new customer deepens the loss.

CAC = total acquisition spend for a period ÷ new customers in the same period.

Acquisition spend includes:

• advertising and paid channels: search, social, affiliate payouts;
• salaries and bonuses of marketing and sales staff, including employer Social Security contributions in Spain;
• sales and marketing tools: CRM, email, analytics;
• contractors: agencies, content, events.

This gives two different CACs. Blended CAC is all acquisition spend divided by all new customers.

Paid CAC is paid-channel spend only, divided by the customers those channels brought in.

Andreessen Horowitz draws this distinction in its “16 Startup Metrics” overview.

An illustrative example — the same subscription service, acquisition spend for one month:

• advertising — €4,500, which brought in 30 customers;
• marketing and sales salaries with taxes and contributions — €3,600;
• CRM, email, analytics and contractors — €900;
• 50 new customers in total: another 20 came on their own, from search and referrals.

The blended CAC of €180 is the same figure used in the calculation above.

Had the founder shown investors €90, LTV/CAC would rise on paper from ≈ 5.1 to ≈ 10.3 — and collapse at the first question about sales salaries.

Paid CAC tells you where to add budget: if it grows faster than customer margin, the channel is running out of steam. Blended CAC is for unit economics and investor conversations.

CAC payback period

CAC payback = CAC ÷ monthly margin per customer. Use margin, not revenue: revenue still has to cover hosting, fees and support.

In the example: 180 ÷ 37 ≈ 4.9 months. On revenue it would be 180 ÷ 49 ≈ 3.7 months — a quarter shorter than reality.

A benchmark investors often quote is CAC payback within 12 months.

David Skok (Matrix Partners) set it out on his For Entrepreneurs blog back in 2011, noting himself that longer periods are acceptable when selling to large companies.

Bessemer Venture Partners, in its “Scaling to $100 Million” research, also ties the benchmark to customer size: up to 12 months for SMB, up to 18 for mid-market and up to 24 for enterprise customers.

LTV/CAC: what ratio is good

LTV/CAC = margin-based LTV ÷ blended CAC. It shows how many euros of margin each euro spent on acquisition brings back.

• Below 1 — every customer loses money: more sales mean bigger losses.
• Around 1–2 — acquisition pays back, but little is left for fixed costs and profit.
• 3 or more — a benchmark investors often quote, popularised by the same David Skok.
• Far above 3 — the model is healthy, but sometimes it signals under-investment in growth: the company could acquire customers more aggressively.

3:1 is a guideline, not a standard. The acceptable ratio depends on margin, churn and cash: with long payback even a good LTV/CAC ties up a lot of working capital. That is why LTV/CAC is read together with payback, not instead of it.

In the example LTV/CAC = 925 ÷ 180 ≈ 5.1 and payback is 4.9 months: both benchmarks are met comfortably.

Common CAC mistakes

• Ad spend only. Without sales and marketing salaries and tools, CAC is understated — by half in the example.
• Organic customers in the denominator. Dividing the ad budget by all new customers, including those from search and referrals, makes paid channels look cheaper than they are.
• Timing mismatch. In B2B, months pass between first touch and payment: January spend brings March customers. Shift the spend period by the length of the sales cycle.
• One average CAC for all channels. A cheap channel hides an expensive one — calculate CAC per channel.
• Payback on revenue. Revenue-based payback is shorter than the real one; use margin.
• Trial users counted as customers. Only paying customers go in the denominator.

MRR and ARR

MRR (monthly recurring revenue) is the sum of payments from active subscriptions, normalised to one month. ARR (annual recurring revenue) is the same figure on an annual basis.

MRR = sum of monthly payments from all active paying customers, excluding VAT.

Annual contracts are divided by 12: a customer who pays €1,200 a year upfront adds €100 to MRR, not €1,200 in the month of payment.

ARR = MRR × 12. ARR is usually reported by companies with annual contracts and larger customers; MRR by services billed monthly.

MRR excludes one-off fees (implementation, training, setup), VAT and free trials. Discounts are taken into account: MRR is what customers actually pay.

What drives the change in MRR

• New MRR — revenue from customers acquired this month;
• Expansion MRR — growth from existing customers: higher plans, extra seats and modules;
• Contraction MRR — decline from existing customers: lower plans, fewer seats;
• Churned MRR — revenue from customers who left.

Net new MRR = New + Expansion − Contraction − Churned. This breakdown shows where growth comes from: acquisition or existing customers.

An illustrative example: at the start of the month the service has 500 customers at €49, so MRR = €24,500.

During the month it gains the same 50 new customers as in the CAC calculation.

Without new customers MRR would have fallen to €24,070 — 98.2% of the starting figure. This share is called net revenue retention (NRR).

Above 100%, existing customers pay more and more, and revenue grows even without new sales.

How ARR differs from annual revenue

ARR is a snapshot at a date: how much recurring revenue current subscriptions will bring over 12 months if nothing changes.

Annual revenue is a fact: what the company actually earned from January to December, including one-off fees.

For a growing service, year-end ARR is well above revenue for the year.

If MRR grew evenly from €12,000 in January to €26,520 in December, subscription revenue for the year is about €231,000, while year-end ARR is €318,240.

The opposite mistake is multiplying your best month, one-off fees included, by 12 and calling it ARR.

In the same “16 Startup Metrics”, Andreessen Horowitz warns that ARR and annual revenue run-rate are not the same thing.

It is safer to show MRR, its components and revenue separately — that is how investors expect them on the traction slide of your pitch deck, and ARR growth is what burn rate and runway efficiency is measured against.

Fixed and variable costs

Unit economics focuses primarily on variable costs — those that depend directly on sales volume or customer numbers:

• cost of the product or service;
• payment processing fees;
• delivery to the customer.

Fixed costs — office rent, administrative salaries, fixed licences — are usually not included directly in basic unit economics; they are analysed separately in the full financial model.

The key idea: unit economics shows whether each unit of the business is profitable on its own.

Then you estimate how many units you must sell to cover fixed costs — 486 customers in the example above.

Where and how unit economics is calculated

• in the project’s financial model — on a separate unit calculation sheet;
• in a unit economics calculator — usually Excel or Google Sheets;
• in the investor pitch deck — one or two slides showing the model is viable.

At the start a simple table is enough: CAC, projected LTV, margin and payback. Even a minimal calculation shows whether to scale the idea or fix the model first.

Unit economics is a compact mirror of the business: is it working in the black or not?

The first real numbers come from testing your niche: test ads reveal both the cost of a customer and their ticket.

Why founders need unit economics before launch

It is not a luxury for after launch; without it any business calculation becomes guesswork. Without unit economics a founder cannot answer basic questions:

• how much marketing money is really needed to acquire the required number of customers;
• when the project will start making a profit rather than just “looking promising”;
• how much a wrong hypothesis will cost and whether there is a safety margin to fix it;
• whether the model pays off at all, or loss-making is built in from day one.

Unit economics reveals the project’s weak spots before real money starts leaking into them: where the risks are, where metrics can be improved and where strategy should be rethought.

Above all, it moves the founder from “hoping for luck” to deliberate management.

Mistakes of founders who ignore unit economics

• Betting on revenue growth without calculating profitability. You can grow sales at a loss. If every deal loses money, more volume only accelerates cash gaps.
• Raising investment without an evidence-based economic model. Investors put money not into a founder’s enthusiasm but into a predictable financial model. Without clear unit economics the project is unattractive or its valuation drops sharply.
• Scaling a loss-making model — “the more we sell, the more we lose”. The startup becomes a machine for accelerating losses.
• Failing to adapt. When prices, customer behaviour or acquisition costs change, the model must change too. Without unit economics founders act blind and keep investing in hypotheses that do not work.
• LTV on revenue rather than margin, and CAC counting only ad spend without sales team salaries. Both make the model look better than it is.

Unit economics is not box-ticking; it is a live barometer of the business, without which you cannot spot risks and change course in time.

Unit economics in investor negotiations

An investor does not buy your idea, emotions or ambitions — they invest in an economic model that can grow and make a profit.

Showing real unit economics is not just an argument in negotiations; it demonstrates the project’s financial maturity and respect for the investor’s money.

A founder who can answer three questions — what a customer costs, what they bring in and how fast they pay back — moves from “dreamer startups” to “startups you can do serious business with”. Investors will look for the same numbers in your traction.

Frequently asked questions

What is unit economics in simple terms?

A calculation of whether the business makes money on one customer or sale: what it costs to acquire a customer, how much margin they bring and how fast they pay back.

How do you calculate LTV?

Margin per customer per period × the average number of periods a customer stays. With constant churn the average number of periods is roughly 1 ÷ churn. Calculating on revenue rather than margin is a common mistake.

What is CAC and how do you calculate it?

CAC is the cost of acquiring one new paying customer. Total acquisition spend for a period — advertising, marketing and sales salaries, tools — is divided by the number of new customers in the same period. Counting ad spend alone understates CAC.

What LTV/CAC ratio is considered good?

A benchmark investors often quote is LTV at least three times CAC and acquisition payback within 12 months, popularised by David Skok. It is a guideline, not a rule: selling to large companies justifies longer payback. LTV should be calculated on margin, not revenue.

What is the difference between MRR and ARR?

MRR is monthly recurring subscription revenue; ARR is the same figure for a year: ARR = MRR × 12. MRR suits services billed monthly, ARR suits annual contracts. ARR is not annual revenue: it is a snapshot of current subscriptions, excluding one-off fees.

Are fixed costs part of unit economics?

Not in the basic calculation: unit economics looks at variable costs. Fixed costs are covered by total margin, and you separately calculate how many customers that takes.

Can you calculate unit economics before launch?

Yes, from hypotheses, then refine with early data: test ads give acquisition cost, first sales give ticket and margin, the first months give churn.

Key points about unit economics

• Unit economics shows whether the business makes money on each customer.
• Core metrics: CAC, margin, LTV, payback, retention and churn.
• Calculate LTV on margin and revenue excluding VAT.
• Use blended CAC: advertising, sales and marketing salaries, tools.
• Net new MRR = New + Expansion − Contraction − Churned; ARR = MRR × 12 and is not annual revenue.
• The investor rule of thumb: LTV/CAC of at least 3 and acquisition payback within a year.
• If unit economics do not work, scaling only accelerates losses.

The subscription example is illustrative; all results are calculated from the stated assumptions.

LTV/CAC and CAC payback benchmarks: David Skok, For Entrepreneurs, “SaaS Metrics 2.0”; Bessemer Venture Partners, “Scaling to $100 Million”. Blended vs paid CAC, ARR vs run-rate: Andreessen Horowitz, “16 Startup Metrics”.