Marketing Cost Per Customer: What the Number Is Telling You

Marketing cost per customer is the total marketing spend required to acquire one paying customer over a given period. You calculate it by dividing your total marketing expenditure by the number of new customers acquired in the same timeframe. Simple arithmetic. But the number is almost never as simple as the formula suggests.

What makes cost per customer genuinely useful is not the headline figure. It is what sits underneath it: which channels are doing real work, which customers are worth acquiring, and whether your marketing is creating demand or just converting it. Most businesses track the metric. Far fewer interrogate it properly.

Why the Formula Is the Easy Part

Total marketing spend divided by new customers acquired. You can do that in ten seconds. The harder question is whether the output means anything useful.

Early in my career I ran performance campaigns with a level of confidence I no longer have. The numbers looked excellent. Cost per acquisition was tight, return on ad spend was strong, and the dashboards were full of green. What I did not fully appreciate then was how much of that performance was simply capturing people who were already going to buy. We were fishing in a stocked pond and congratulating ourselves on the catch.

The problem with blended cost per customer is that it averages across very different types of acquisition. A branded search conversion, where someone already knew the company and typed its name into Google, sits in the same bucket as a cold display campaign that reached someone who had never heard of the brand. One of those is harvesting intent. The other is creating it. They have completely different economics, completely different strategic value, and they should never be averaged into a single figure without qualification.

If you are working through how cost per customer fits into a broader commercial strategy, the articles in the Go-To-Market and Growth Strategy hub cover the surrounding decisions: channel selection, audience development, and how to build acquisition models that hold up under scrutiny.

How to Calculate Marketing Cost Per Customer Properly

The base calculation is straightforward. If you spent £500,000 on marketing in a quarter and acquired 2,000 new customers, your marketing cost per customer is £250. That is the starting point, not the answer.

To make the number meaningful, you need to make several deliberate choices about what goes into each side of the equation.

What counts as marketing spend

Paid media is obvious. But what about the salaries of your in-house team? Agency fees? The cost of content production that supports organic acquisition? Technology and tooling? If you include some of these and not others, your cost per customer becomes incomparable across time and incomparable against any external benchmark. Be consistent and be explicit about what is in the number.

I have seen businesses report a cost per customer that excluded their entire marketing headcount because those people sat on a different cost centre. The number looked efficient. The actual economics were not. When we rebuilt the model to include the full cost of running the function, the picture changed considerably. Not catastrophically, but enough to shift decisions about channel mix and team structure.

What counts as a new customer

This sounds obvious until you look closely. Are you counting first purchase? First subscription? First qualified lead that converts? Are you including reactivated lapsed customers? If someone bought from you three years ago and comes back, are they new or returning? None of these definitions is wrong, but you need one definition applied consistently.

For businesses with longer sales cycles, there is also the timing problem. The customer acquired in Q3 may have been touched by marketing spend in Q1 and Q2. A simple quarterly calculation will attribute the acquisition to the wrong period. Cohort-based analysis, where you track spend and acquisition together for a defined group over time, gives you a more honest picture.

The Attribution Problem That Inflates Your Numbers

Marketing attribution is one of the most consequential and least honest conversations in the industry. Every channel wants credit. Most attribution models give it to them.

Last-click attribution, still the default in many businesses, assigns the entire value of a conversion to the final touchpoint before purchase. This systematically rewards lower-funnel channels, particularly branded search and retargeting, and systematically undervalues the channels that built awareness and intent in the first place. The result is a cost per customer calculation that makes performance channels look brilliant and brand investment look wasteful.

When I was running a large performance marketing operation, we had a client who was convinced that their paid search programme was the engine of their business. The cost per customer from that channel was genuinely low. What we eventually demonstrated, through a series of geo-holdout tests, was that a significant portion of those conversions were people who had been reached by TV and out-of-home advertising weeks earlier. Switch off the brand spend, and the search volume dropped. The performance channel was not generating demand. It was the final step in a experience that started somewhere else entirely.

This is not an argument against performance marketing. It is an argument for being honest about what it is doing. Forrester’s thinking on intelligent growth has long pointed toward the need for businesses to understand the full demand generation picture, not just the final conversion event. The channel that closes the sale is not always the channel that made the sale possible.

Cost Per Customer Versus Customer Acquisition Cost

You will often see marketing cost per customer used interchangeably with customer acquisition cost, or CAC. They are related but not identical, and the distinction matters.

CAC typically includes sales costs alongside marketing costs. For businesses with a direct sales function, particularly in B2B, the sales team’s salaries, commissions, and tools can dwarf the marketing spend. A SaaS business with a large enterprise sales team might have a CAC of £8,000 even if the marketing cost per customer is only £1,200. Conflating the two leads to poor decisions about where to invest and where to cut.

Marketing cost per customer is a narrower, more controllable metric. It tells you what the marketing function is spending to bring customers to the door. CAC tells you what it costs the whole business to close them. Both matter. They answer different questions.

For product-led growth models, where the product itself is the primary acquisition mechanism, the relationship between these numbers looks different again. Growth models built around product virality and referral loops can produce very low marketing cost per customer figures, but only because the product is doing work that marketing would otherwise have to pay for. The cost has not disappeared. It has shifted into product development and retention investment.

The Relationship Between Cost Per Customer and Lifetime Value

A cost per customer figure without a lifetime value figure is almost meaningless for strategic decisions. Acquiring a customer for £50 is excellent if they spend £2,000 over three years. It is a disaster if they churn after one purchase worth £40.

The LTV:CAC ratio is the metric that actually tells you whether your acquisition economics are sustainable. A ratio of 3:1 is commonly cited as a reasonable benchmark for subscription businesses, meaning you recover the cost of acquisition three times over in customer lifetime value. But that benchmark comes from SaaS contexts and does not translate cleanly to retail, services, or non-subscription models. Apply it with caution.

What the ratio does do is force a conversation about customer quality, not just customer volume. I have worked with businesses that were proud of their falling cost per customer, only to find that the customers they were acquiring at lower cost had significantly shorter lifespans and lower average order values. They had optimised for the wrong thing. The acquisition machine was efficient. The business was not growing.

This is where the tension between short-term efficiency and long-term growth becomes visible. Optimising hard for cost per customer tends to push you toward audiences already close to purchase, messaging that converts quickly, and channels with tight feedback loops. All of that is rational in isolation. The problem is that it systematically neglects the work of reaching people who do not yet know they need you, which is where most future growth actually comes from.

Why Low Cost Per Customer Can Be a Warning Sign

There is a version of low cost per customer that should worry you. If your number is falling consistently and you are not doing anything meaningfully different, one of several things may be happening.

First, you may be exhausting your addressable audience. When a brand has strong awareness in a defined market, the early cohorts of customers are relatively cheap to acquire because they are already warm. As you work through that pool, you either need to expand into new audiences, which costs more, or accept that the cheap customers are gone and the next ones will be harder to reach. A falling cost per customer in this scenario is not efficiency. It is the last easy wins before the curve steepens.

Second, you may be over-indexing on channels that are harvesting rather than building. Retargeting, branded search, and email to existing prospects are all efficient on a cost per customer basis. They are also all dependent on a stock of interested people that was created by other activity. If you are cutting brand spend to fund performance spend because performance looks cheaper, you are borrowing against future demand without acknowledging the debt.

Third, you may simply be measuring it wrong. Attribution models that favour last-touch will make performance channels look increasingly efficient over time, even as the underlying economics deteriorate. I have seen this pattern more than once. The dashboard looks better every quarter. Then a competitor enters the market, or a platform changes its algorithm, and the cost per customer spikes sharply because the brand had no residual awareness to fall back on.

The BCG work on go-to-market strategy has long argued that sustainable commercial models require investment across the full customer development curve, not just at the point of conversion. Cost per customer metrics that only measure the final step miss most of what makes acquisition possible in the first place.

How to Segment Cost Per Customer for Better Decisions

The aggregate cost per customer figure is a management summary. The segmented version is where you find the decisions.

Break it down by channel first. What does it cost to acquire a customer through paid social versus organic search versus referral versus direct mail? These numbers will be very different, and the differences are informative. But do not stop at the channel level. Ask what those customers do after acquisition. A customer acquired through referral might cost three times as much as one acquired through paid search, but if they retain at twice the rate and spend 40% more, the referral channel is the better investment despite the higher upfront cost.

Segment by product or offer type. If you run promotions to drive acquisition, the cost per customer during a promotional period will look artificially low. Those customers may be deal-seekers with poor retention profiles. Mixing them into your standard cost per customer figure distorts the baseline.

Segment by geography if you operate across multiple markets. Cost per customer varies enormously by market maturity, competitive intensity, and brand awareness. A business expanding into a new region should expect a significantly higher cost per customer than in its home market. That is not a failure. It is the cost of building a new customer base from a standing start. Benchmarking the new market against the established one is a category error that kills expansion investments before they have time to work.

Segmenting by acquisition cohort over time also reveals something important. If customers acquired in year one have a different long-term value profile than customers acquired in year three, that tells you something about how your acquisition mix has shifted and whether the quality of what you are bringing in is changing. Understanding how growth loops affect retention and referral is part of reading these cohort patterns honestly.

The Cost of Acquiring Customers You Did Not Need to Market To

One of the most uncomfortable truths in acquisition marketing is that a meaningful proportion of the customers you are paying to acquire would have found you anyway. They were already looking. They already had intent. Your ad was the last step, but it was not the cause.

This matters for cost per customer because it means part of your marketing spend is not generating incremental customers. It is accelerating or confirming decisions that were already in motion. That is not entirely without value. Speed and convenience matter. But it is a different kind of value than reaching someone cold and building a relationship from nothing.

The clothing shop analogy has stayed with me for years. Someone who tries something on is many times more likely to buy than someone browsing the rail. The fitting room does not create the purchase intent. But the shop that has no fitting room loses customers who needed that final step. Some of your marketing spend is the fitting room. It is necessary, but it is not the reason people wanted the clothes.

Incrementality testing, where you hold back a control group from seeing your marketing and measure the difference in conversion rates, is the most honest way to understand how much of your cost per customer is buying real new customers versus paying for conversions that were already going to happen. It is not easy to run, and it requires a willingness to see numbers that may be uncomfortable. But it is the only way to know what your acquisition spend is genuinely worth.

The Vidyard research on untapped pipeline potential points to a consistent pattern: GTM teams systematically underestimate how much demand is sitting unaddressed because they are too focused on converting the demand they can already see. Cost per customer metrics, when tracked only at the bottom of the funnel, reinforce exactly this blind spot.

When Marketing Cost Per Customer Is Not the Right Metric

There are situations where cost per customer is the wrong thing to optimise for, and treating it as the primary success metric will lead you somewhere you do not want to go.

Brand-building investment does not produce customers in the short term. If you run a campaign designed to shift perception, build category relevance, or reach people who are 12 months from a purchase decision, measuring it on cost per customer will make it look like a failure. The metric is simply not designed to capture that kind of work. Using it to evaluate brand investment is like judging a restaurant on how quickly the food arrives without asking whether it tastes good.

Early-stage businesses in new categories face a similar problem. If you are creating a market rather than competing in one, the cost of educating the audience and building the category is real and necessary, but it will not show up as efficient customer acquisition in the short term. Applying a cost per customer lens too early can kill category creation before it has a chance to compound.

Retention and expansion marketing also get distorted by this metric. The cost of keeping a customer, upselling them, or turning them into an advocate is marketing spend, but it does not produce new customers in the way the formula assumes. If you are running a sophisticated customer marketing programme and folding that spend into your cost per customer calculation, you are making the number worse without making the business worse. Separate the two budgets and measure them separately.

There is a broader point here about the relationship between marketing and business health. I have spent time working with companies that were using marketing spend to compensate for product problems, service failures, and pricing that did not hold up in the market. The cost per customer kept rising not because marketing was inefficient, but because the business was not giving customers a compelling reason to come back. No amount of acquisition optimisation fixes a leaky bucket. If retention is poor, the cost per customer calculation will always be fighting against the tide.

The Forrester analysis of go-to-market struggles in complex markets identifies a recurring pattern: organisations that focus acquisition metrics on marketing efficiency while underinvesting in the customer experience end up in a cycle of rising acquisition costs and falling retention. The metric becomes a symptom of a deeper problem rather than a lever for solving it.

Building a Cost Per Customer Model That Holds Up

If you want a cost per customer model that is actually useful for decisions, it needs a few things that most businesses do not have in place.

A consistent definition of what is included in marketing spend, applied across every period you are measuring. A consistent definition of what counts as a new customer. A channel-level breakdown so you can see where the efficiency is coming from. A lifetime value figure attached to each channel segment so you can compare like for like. And some form of incrementality validation so you know how much of the spend is genuinely driving new customers versus capturing existing intent.

None of this is technically difficult. Most businesses have the data to do it. What they lack is the discipline to define the terms clearly and the willingness to look at numbers that may complicate the narrative. The cleanest cost per customer figure is often the one that has been most carefully constructed to look clean.

When I was running agency teams, we had a rule about client reporting: if a number looked too good, we interrogated it before we presented it. Not because we wanted to deliver bad news, but because a number that looked too good was usually a sign that something in the methodology was off. That instinct has served me well across a lot of different businesses and a lot of different markets.

The businesses that use cost per customer well treat it as one input into a broader picture, not as the verdict on whether marketing is working. They segment it, stress-test it, and read it alongside retention and lifetime value data. They also accept that some of the most important marketing work they can do, reaching new audiences, building brand relevance, creating demand rather than just capturing it, will make the headline cost per customer figure look worse in the short term before it makes the business look better in the long term.

If you want to go deeper on how acquisition metrics connect to broader growth strategy, the Go-To-Market and Growth Strategy hub covers the full picture: from channel economics and audience development to how growth models actually hold up when you stress-test them against real commercial constraints.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions