How do I calculate what a new customer really costs?
Divide the acquisition costs you included by the new paying customers they produced in the same window. A €1,200 campaign with 40 enquiries, 12 qualified leads and 4 new customers costs €30 per enquiry, €100 per qualified lead and €300 per customer. Only the last one is customer acquisition cost.
A cheap enquiry is not yet a cheap customer. The same campaign can produce three different cost figures. This example shows what each figure buys you, so you do not judge a campaign by the wrong number.
Name what you paid for and what you counted
Customer acquisition cost divides the included acquisition costs by new customers acquired. Enquiries and qualified opportunities are earlier outcomes. They require different labels. State the customer rule too: a first-time paying customer is different from a free registration or an existing customer placing another order.
Include the relevant acquisition work under a consistent cost rule. Make shared-cost allocations and omissions, such as unrecorded owner time, visible. Costs and buying decisions can fall in different months, so check that the comparison gives the activity enough time to produce its observed outcomes. A partial advertising-only figure should not be presented as the full cost of acquisition.
One training campaign, three honest numbers
A fictional training provider spends €600 on advertising, €300 on acquisition-related sales work, €200 on creative work and €100 on its allocated share of tools. Its stated campaign cost is €1,200.
Under its recorded source rule, the campaign has 40 unique enquiries. Twelve meet its written qualification rule. By 31 May, four of those twelve have become first-time paying customers. All four are included in the twelve, and all twelve are included in the forty.
The same €1,200 works out at €30 per enquiry, €100 per qualified opportunity or €300 per new customer. The spending has not changed. The group underneath the division has changed.
That is the point: ask what happened before deciding what the number means. Calling €30 the cost of a customer would make this campaign appear ten times cheaper at acquiring customers than its observed result. Whether €300 is worthwhile still depends on the value and costs of the customer relationships.
- The campaign and values are fictional. All listed acquisition costs are allocated to this campaign under the stated rule.
- The 40 enquiry records are unique; the 12 qualified opportunities contain the four new paying customers. Results are observed by 31 May.
- Qualification follows the provider's written rule. It is not a universal definition or a promise to buy. Delivery costs and customer returns are not supplied.
| Measure | Calculation | What it describes |
|---|---|---|
| Cost per enquiry | €1,200 ÷ 40 = €30 | An enquiry received |
| Cost per qualified opportunity | €1,200 ÷ 12 = €100 | An enquiry meeting the written qualification rule |
| Cost per new paying customer | €1,200 ÷ 4 = €300 | A first-time customer who has paid |
Try it with your own numbers
What the campaign's adverts cost in the period.
The cost of time spent winning these customers. If you leave it out, say so next to the result.
Design, writing or other work made for this campaign.
Enquiries from this campaign, each counted once.
Enquiries that meet your written qualification rule.
First-time customers who have paid, counted by the same date.
Your result
Fill in every field to see the result.
Your numbers stay in this browser. Nothing you type is sent or saved.
One customer can encounter several sources
Attribution means assigning credit to contacts that preceded an outcome. Google Analytics describes it as a rule or model for sharing that credit. It is distinct from the count of actual customers.
In the training example, one buyer hears a friend's recommendation, clicks the advert and later searches for the provider before paying. That is one customer with three known contacts. If two reports each claim the purchase, adding their customer totals counts the same buyer twice. A first-contact or last-contact rule can organise a comparison; it does not prove that removing all the other contacts would leave the purchase unchanged.
Cost per outcome is only one part of the decision
The €300 in the example describes acquisition spending per observed new customer. The example gives no purchase amount or delivery cost, so it cannot show whether those customers repay that spending. Nor does a €30 enquiry price tell us whether an enquiry is useful.
If another campaign has no qualified opportunities, dividing its spending by zero does not produce a zero cost per opportunity. Report the spending and the zero outcomes separately. If some decisions are pending, show them. More customers may arrive later, but that is an unresolved possibility, not income already earned.
Common misconceptions
- A qualified opportunity is an acquired customer.
- Qualification establishes that an opportunity meets a stated rule. The person may still decline or remain undecided.
- Two channels claim a purchase, so there were two purchases.
- Check the underlying customer and transaction records. Reports can assign credit differently to the same outcome.
- The lowest enquiry cost identifies the best campaign.
- This comparison does not yet contain customer outcomes, delivery costs or the value of the resulting relationships.
Make one ambiguous cost figure usable
Write its full label: which costs, divided by which outcomes, from which activity, checked on what date. Note pending decisions and omitted costs. Trace one purchase across its recorded sources to check for duplicate credit. Then compare like measures; do not change a budget just because one report uses a smaller denominator or a different label.
Sources
- Stripe: CAC in SaaS (opens in a new tab)The new-customer denominator, included acquisition costs, shared-cost allocation and timing across periods. SaaS benchmarks and revenue-only payback claims are not adopted.
- Google Analytics: Get started with attribution (opens in a new tab)Attribution assigns credit across contacts preceding an outcome. The fictional contact sequence illustrates why credited outcomes need deduplication.