Unit Economics: How to Work Out What Each Customer Earns You

What Is Unit Economics?

Unit economics is the profit or loss a business makes on one unit of what it sells, after the costs tied directly to that unit. For most B2B companies the unit is one customer: what they pay you over the whole relationship, minus what it costs to serve them and what it cost to win them. A project firm can use one engagement, and a shop can use one order.

The calculation rests on three numbers: the profit one customer brings in each year, how many years they stay, and the full cost of winning them. Get those three right and you can answer the question that matters before you grow: does winning one more customer make you money or cost you money?

How to Calculate Unit Economics in 5 Steps

You need a year of accounts, your customer list and an afternoon. Use the last twelve months, so seasonal swings even out. The figure shows how the lines in your books turn into the two checks at the end.

1. Pick the Unit

Choose the thing you sell that repeats. For a firm with retained or recurring clients, that is one customer. For a firm that sells one-off projects to different buyers, it is one project. For a shop, it is one order.

Then split your customers into groups if they differ a lot. A firm with a few large accounts and many small ones has two businesses inside it, and one average hides both. Work the numbers for each group. The blended figure comes last, if at all.

Do this: write down the unit and the two or three customer groups you will compare.

2. Work Out Profit per Customer

Profit per customer per year = average revenue per customer minus the cost to serve one customer.

The cost to serve is every cost that rises when you add a customer and falls when you lose one. In a services firm, that is the time of the people who deliver the work at what they cost you, plus subcontractors, travel and any software you pay for per client. In a product or software company, it is hosting, support, payment fees and onboarding. Leave out rent, leadership salaries and finance. Those are overheads, and they belong to the whole company, not to one customer.

Services firms most often get the delivery cost wrong. Count each person's full cost, salary plus benefits and payroll taxes, and remember that how many hours each consultant bills sets what a client really costs: unbilled hours are still paid for, so spread them across the clients.

Do this: take revenue by customer group from your accounts, subtract the cost to serve each group, and write the result as a dollar figure and as a share of revenue.

3. Work Out the Full Cost to Win a Customer

Cost to win one customer = all sales and marketing cost in the period, divided by the new customers won in the same period.

Most guides count ad spend and stop there. That is the flattering version. Count everything you spend to win business: salespeople's salaries, commissions and benefits, marketing staff, agencies, events, content, software and travel.

Then add the founder's own selling time. If you spend a third of your week on prospects and proposals, a third of what it would cost to hire someone into your role belongs here. In a company where the founder still closes most deals, that time is often the largest cost to win, and the one nobody writes down. When you are deciding how much of the budget goes to winning new customers, this is the number that tells you whether the money is coming back.

Do this: total last year's sales and marketing costs, add your own selling time at a replacement salary, and divide by the number of new customers won.

4. Work Out Lifetime Value

Lifetime value = profit per customer per year times the number of years a customer stays.

Find how long customers stay from your own records. Take the customers you had three years ago and count how many are still with you. If one in four leaves each year, the average customer stays about four years, because the years a customer stays equal 1 divided by the share who leave each year.

Cap the answer at five years, even if your customers stay longer. A lifetime value that depends on year eight is a guess, and a guess is a poor reason to spend money now. Use profit, not revenue. In the example below, a lifetime value built on revenue would look almost three times larger and say nothing about cash.

Do this: work out the share of customers you lose each year, turn it into years, and multiply by yearly profit per customer.

5. Run the Two Checks

Check one: lifetime value divided by the cost to win. This tells you whether a customer is worth winning at all. Below 1, every new customer loses money. Around 3 is the usual guideline.

Check two: months to earn back the cost to win = the cost to win divided by the monthly profit per customer. This tells you how long your cash is tied up. The longer it takes, the more cash each new customer uses before it pays anything back, and the faster growth drains the bank.

A business can pass the first check and fail the second: customers worth winning, but so slow to pay back that growth runs out of cash first.

Do this: put both numbers on one page for each customer group, and work them out again every quarter.

Unit Economics Example: A Services Firm

Take a B2B consulting firm with 30 clients paying an average of $120,000 a year, or $3.6 million of revenue. It wins about ten new clients a year and loses about as many. The figures are made up to show the method; they are not from a real company.

The founder's $75,000 is 30 percent of the $250,000 it would cost to hire someone into the role. The 36 percent margin is ordinary: the average firm in SPI Research's 2025 benchmark of 403 professional services firms kept a project margin of 35.9 percent in 2024.

Now count marketing alone, the way most guides do. The cost to win falls to $12,000 and the ratio jumps to 10.8. On that number, the founder would hire two more salespeople tomorrow. On the full number, the firm is healthy but not ready to pour money into sales. It needs to keep clients longer or earn more from each one first.

The Same Steps for a Software or Product Company

The five steps are the same, and two inputs change. Profit per customer comes from gross margin, which is high in software: B2B software companies held gross margin at about 80 percent in 2025, according to Benchmarkit and Aleph's 2026 survey of 342 companies. How long a customer stays comes from the share of revenue you keep each year. In the same survey, the median company kept 84 percent of the revenue from its existing customers in 2025, before any upsell.

Take a subscription at $24,000 a year with an 80 percent margin. It earns $19,200 a year, or $1,600 a month. If it costs $32,000 to win the customer, it takes 20 months to earn that back. Keeping 84 percent of revenue a year implies a customer stays a little over six years. Capped at five, lifetime value is $96,000 and the ratio is 3.0.

The cost to win depends most on whether the product or a salesperson wins the customer: a product bought without a sales call costs far less to win than one that needs a sales team.

What Good Looks Like: The Benchmarks

Good unit economics means a customer earns back what it cost to win them quickly, then earns several times that before they leave. The two guidelines most people quote come from the investor David Skok: a lifetime value more than 3 times the cost to win, and a payback under 12 months. He calls them "only guidelines", and he wrote them for software companies.

Two things matter more than any benchmark. The first is direction: a ratio sliding from 4 to 3 over four quarters is a warning, even though 3 passes. The second is the spread between groups: a firm averaging 3.5 may have one group at 6 and another below 1, and the second group is where the cash goes.

Unit Economics vs Gross Margin, Contribution Margin and EBITDA

All four measure profit. They differ in what they subtract, and in whether they look at one customer or the whole company.

A company needs all four. Unit economics is the only one that tells you whether to spend more on growth.

What to Fix First When the Numbers Are Bad

Fix them in this order. Each fix is usually faster and cheaper than the one after it.

1. Raise Prices

Price goes straight to profit per customer, with no new cost. In the example, an 8 percent rise takes revenue per client from $120,000 to $129,600 while the cost to serve stays at $76,800. Profit per client rises 22 percent, from $43,200 to $52,800. The ratio moves from 3.5 to 4.2, and payback falls from 10.4 months to 8.5. Most firms can raise prices without losing the customers they want by starting with new clients and with the accounts that take the most work for the least money.

Do this: price every new proposal at the higher rate for one quarter and track how many you win.

2. Cut the Cost to Serve

Standardize how the work is delivered, so each client takes fewer hours. In SPI's benchmark, the most mature services firms kept project margins of 55.8 percent against 16.5 percent in the least mature. The survey shows the two move together, not that one causes the other.

Do this: list the five tasks your team repeats for every client and write down one standard way to do each.

3. Keep Customers Longer

An extra year of a client's life adds a full year of profit at no new cost to win. In the example, keeping clients four years instead of three lifts lifetime value from $129,600 to $172,800 and the ratio from 3.5 to 4.6. Keeping the customers you already have starts with knowing why the last ones left.

Do this: call every client who left in the last year and ask what would have made them stay.

4. Cut the Cost to Win

Split your cost to win by channel. Referrals, events, outbound and paid ads rarely cost the same, and they rarely bring in customers who stay as long. Cut the channel with the worst ratio of lifetime value to cost, not the one with the biggest bill.

Do this: tag every customer won last year with the channel that brought them in, and run the two checks for each channel.

5. Stop Selling to the Customers Who Lose You Money

If one group still fails both checks after the first four fixes, stop selling to it. Losing that revenue feels like going backward, but it frees the people and cash the group was using up. Watch that what is left does not end up with too much revenue resting on a few clients.

Do this: name the group, stop new sales to it this quarter, and decide by each renewal whether to reprice the account or let it go.

Why Buyers and Investors Check These Numbers First

A buyer pays for profit that will repeat after you leave, and unit economics is how they test whether it will. Profit per customer shows whether growth will add profit or only revenue. How long customers stay shows how much of next year's revenue is already in hand. The full cost to win, including the founder's time, shows what the business will cost to run once you stop selling.

Expect a buyer to rebuild these numbers customer by customer in due diligence, because they sit close to the heart of what a private equity buyer looks for. Work them out before a buyer does, and you get to fix them first.

Frequently Asked Questions

What is a good LTV to CAC ratio?

Above 3 is the usual guideline: a customer's lifetime profit should be more than three times what it cost to win them. Below 1, each new customer loses money. A very high ratio can mean you are spending too little on growth. Treat 3 as a guideline, not a rule, and watch the trend over several quarters.

Can a company have good unit economics and still lose money?

Yes. Unit economics leaves out overheads such as rent, leadership salaries and finance. A company can make money on every customer and still lose money overall if it has too few customers to cover those fixed costs. Growth fixes that kind of loss. It makes the opposite kind, losing money on each customer, worse.

What does negative unit economics mean?

It means a customer earns less over their life than it cost to win and serve them, so lifetime value divided by the cost to win is below 1. Every new customer makes the loss bigger. The fix is price, the cost to serve, how long customers stay or the cost to win, not more sales.

How often should I check my unit economics?

Every quarter, on the last twelve months of numbers, split by customer group. Quarterly is often enough to catch a trend before it costs much, and twelve months smooths out seasonal swings. Work them out again straight away after a price change, a new sales hire or a new marketing channel.

Should the founder's time count in the cost to win a customer?

Yes. Value it at what you would pay someone to do the selling you do now. Leaving it out makes the cost to win look lower than it is, and the true figure appears the day the founder stops selling and a salesperson has to be hired.

Final Thought

Unit economics will not tell you how to grow. It tells you whether growing will make the company richer or poorer, and which lever to pull first: price, the cost to serve, how long customers stay or the cost to win them. That is the N of the UNLOCK Method: name the growth levers. If your revenue is climbing and your bank balance is not, Mark's work as a growth advisor starts by finding which lever to pull first, then stays through the work of pulling it.