Here’s a question worth being able to answer quickly: when you win one more customer, does the business get stronger or weaker?

Most founders assume stronger, and often it is. But the costs of winning and serving a customer are spread across several lines of your profit and loss, none of them labelled, which means the answer is rarely as obvious as it feels. Plenty of startups have scaled hard on the assumption it was positive and found out otherwise.

In this blog, we’ll explain how to work out your unit economics – what a customer really costs you, which number to lead with when you’re testing it, and what the answer tells you about when to spend harder on growth.

Picking the unit

For most startups the unit is a customer. For marketplaces it’s usually a transaction or an active seller; for usage-based products it might be an account rather than a user.

What matters is that the unit is something you can count and something you spend money to acquire. Get this wrong and every number downstream is measuring a different thing each quarter.

Where you have genuinely different customer types – a self-serve tier and an enterprise tier, say – calculate them separately. Blended unit economics across two very different motions produce a number that describes neither.

Fully loaded acquisition cost

Customer acquisition cost is total spend on winning customers, divided by customers won in the same period. The important word there is “total”.

It includes paid media, content and marketing tooling, the salaries and commission of everyone in sales and marketing, agency fees, event costs and the share of any founder time genuinely spent selling. Not just the ad spend, which is the version most dashboards show.

Two refinements worth making. Split paid and organic acquisition, because they scale differently and blending them hides the fact that your paid channel is expensive. And lag the calculation to your sales cycle – if deals take three months to close, this month’s cost won this month’s customers only in the loosest sense.

Our article on cutting customer acquisition costs covers what to do once the number is uncomfortable.

Contribution margin, not gross margin

Gross margin is revenue less cost of sales. Contribution margin goes further: revenue from a customer, less everything it costs to serve that specific customer.

For a software business that means hosting attributable to them, payment processing, support time, customer success time, third-party licences consumed on their behalf, and any implementation or onboarding cost.

The onboarding cost is the one most often missed. A customer who needs two days of setup before they generate a penny has a very different profile from one who self-serves, even at the same contract value.

What you want is contribution margin per customer per month. That’s the number that funds everything else.

Why payback period is the more useful number early on

Lifetime value has a credibility problem at early stage: it depends on a churn rate measured over a short period on a small number of customers, and small errors in churn become large errors in LTV.

Payback period avoids that problem. It’s acquisition cost divided by monthly contribution margin, and it tells you how many months a customer takes to repay what they cost to win.

It’s a cash question as much as a profitability one. A twenty-month payback means every new customer is a twenty-month hole in your cash flow before it becomes a contribution – which is why fast-growing companies with good margins can still run out of money.

As a rough guide, under twelve months is strong for a startup, twelve to eighteen is workable, and beyond two years you need either a very low churn rate or a very patient investor.

A worked example

A B2B software startup charges £800 a month. Gross margin looks like 85%, so it feels healthy.

Fully loaded acquisition: £11,000 a month on paid channels, two salespeople costing £13,000 a month including employer costs, and £2,000 of tooling. That’s £26,000 to win eight customers – £3,250 each.

Contribution: of the £800, hosting and processing take £120, and support and customer success average £150 a month per account. Onboarding costs a one-off £400. Contribution margin is £530 a month, not the £680 that gross margin implied.

Payback is £3,650 including onboarding, divided by £530, which is just under seven months. That’s a business that can safely spend more on acquisition. Change one thing – move support cost to £350 a month for a heavier enterprise segment – and payback goes past fourteen months, and the same growth plan starts consuming cash faster than it creates it.

What unit economics decide

They tell you when it’s safe to spend harder on growth. Scaling a business with broken unit economics doesn’t fix them, it multiplies them, and that’s behind most of the margin problems we see in The Silent Killers of Startup Margins.

They also decide how investors read your growth. Revenue growth with deteriorating unit economics looks like buying customers. The same growth with a shortening payback period looks like something that genuinely works, and that difference shows up in how a round goes.

The takeaway

Calculate acquisition fully loaded, use contribution rather than gross margin, and lead with payback period rather than lifetime value. If payback is short and stable, spend more. If it’s lengthening, the problem is the model rather than the growth rate.

Know what your growth actually costs

Disclaimer: This article is general in nature and does not constitute financial advice. Benchmarks vary considerably by business model and sector – speak with a qualified adviser (hey, that’s us!) about what applies to your situation.