Acquisition cost is the most quoted metric in growth discussion and among the most inconsistently calculated, which makes comparisons between businesses largely meaningless.
The basic calculation
Total cost of acquiring customers in a period, divided by customers acquired in that period.
Which is simple and where every dispute begins, since what counts as cost varies.
What should be included
Advertising spend, obviously.
Marketing salaries and contractor costs, since those people are doing acquisition work.
Sales salaries and commission where sales is part of acquisition.
Tools and software supporting those functions.
Agency fees.
Content production costs.
Which produces a figure considerably higher than advertising spend alone, and it is the honest one.
What gets excluded to flatter the number
Salaries, most commonly, on the reasoning that they would be paid anyway.
Which is true and irrelevant — if those people are acquiring customers, their cost is acquisition cost.
Organic acquisition costs, treating anything not from paid advertising as free, which ignores the content and product work that produced it.
Brand spend, on the reasoning that it is not directly attributable, which understates the total.
The attribution period problem
Costs incurred this month may acquire customers next quarter, particularly for long sales cycles.
Which means dividing this month's cost by this month's customers is wrong whenever the cycle is longer than the period.
Cohort-based calculation, tracking spend and the customers it eventually produced, is more accurate and requires tracking that many businesses lack.
Blended and paid
Blended cost divides all acquisition spend by all customers including those acquired organically.
Paid cost divides paid spend by customers attributable to paid channels.
Both are useful and they answer different questions — blended describes the business, paid describes the efficiency of the channel.
Quoting blended cost while implying it describes paid efficiency is a common misrepresentation.
The lifetime value comparison
The ratio everyone quotes and few calculate carefully.
Lifetime value should use gross margin rather than revenue, since acquiring a customer who generates revenue at zero margin generates nothing.
It should be discounted where value arrives over years, since money later is worth less.
And it depends on a retention assumption that is frequently optimistic, particularly for young businesses without enough history to know.
Ratios calculated on revenue and undiscounted lifetime overstate the position substantially.
Payback period
Arguably the more useful metric, and the less quoted.
How long until a customer's gross margin repays their acquisition cost.
Which determines how much cash growth consumes, and it is the number that matters when funding is constrained.
A business with an excellent lifetime value ratio and a long payback period will run out of cash while growing successfully.
The practical version
Calculate honestly, including salaries.
Use gross margin and a defensible retention assumption.
Track payback period alongside the ratio.
And segment by channel and by customer type, since aggregate figures conceal channels that are unprofitable and customers who are.
Marginal cost
An important distinction for decisions about increasing spend.
Average acquisition cost describes what has happened. Marginal cost describes what the next customer will cost.
Which matters because channels saturate — the cheapest customers are acquired first, and cost rises as reach extends.
A business with an acceptable average cost may find the marginal cost of growth well above the level that makes sense, and it will only discover this by measuring the increment rather than the average.
Organic acquisition
Frequently treated as free and never actually free.
Content, product work that drives referral, community management and search optimisation all consume resource.
Which means comparing organic and paid channels honestly requires costing the organic effort, and doing so usually shows it favourably while removing the illusion that it is costless.
Segmenting
Aggregate cost conceals enormous variation between customer types.
Calculating separately by segment frequently reveals that one segment is subsidising another, and that growth in the wrong segment is destroying value while appearing to be progress.
Sales cycle effects
Long cycles mean spend and revenue appear in different periods, which distorts any monthly calculation.
Which is why businesses with long cycles should calculate on a cohort basis, tracking spend to the customers it eventually produced.
Doing otherwise produces figures that swing wildly and mean nothing.
Discounting and its effect
Discounts to win customers are effectively acquisition spend and are rarely counted as such.
Which understates the true cost, sometimes substantially, particularly where introductory pricing runs for a long period.
Including the value of discounts given in the acquisition cost calculation produces a considerably more honest number and is almost never done.
Benchmarks
Published acquisition cost benchmarks are calculated inconsistently across the businesses reporting them, which makes comparison unreliable.
Comparing your own figure across periods, calculated the same way each time, is considerably more informative.
Which also removes the temptation to change the calculation method when the number becomes uncomfortable, a practice that is more common than anyone admits.