SaaS Cost of Sales: What’s Eating Your Margin

SaaS cost of sales is the total spend required to acquire and retain revenue, covering sales headcount, commissions, marketing spend, and the infrastructure that supports both. Most SaaS businesses track it, but fewer understand where it quietly bleeds margin before the numbers become impossible to ignore.

The problem is rarely a single line item. It is the accumulation of small inefficiencies, misaligned incentives, and go-to-market assumptions that made sense at an earlier stage but were never revisited. By the time cost of sales shows up as a board-level concern, the structural causes are usually months old.

What Does SaaS Cost of Sales Actually Include?

The definition varies by company, which is part of the problem. Some businesses include only direct sales costs. Others fold in marketing. A few include customer success. The lack of a consistent definition makes benchmarking almost meaningless unless you are comparing like for like.

A working definition for most SaaS businesses should include: sales team salaries and on-target earnings, commissions and bonuses tied to closed revenue, sales development representative costs, marketing spend attributable to pipeline generation, sales tools and CRM costs, and the portion of customer success that is tied to renewal and expansion rather than pure service delivery. Some CFOs also include a share of product costs where the product itself is a core part of the sales motion, which is common in product-led growth models.

Where it gets messy is the boundary between sales and marketing. In many SaaS businesses, marketing owns demand generation but sales owns pipeline. The cost of creating a lead and the cost of converting it sit in different budgets, owned by different people, measured by different teams. That structural separation makes it easy for cost to accumulate in the gaps without anyone being accountable for the total.

I have seen this play out in businesses that were genuinely confused about why their cost per acquisition kept rising. When we mapped the full experience from first touch to closed deal and included every cost centre that touched that process, the number was routinely 30 to 40 percent higher than the figure the leadership team had been working with. Not because anyone was being dishonest. Because the question had never been asked with enough precision.

Why Blended CAC Is a Comfortable Lie

Customer acquisition cost is the most commonly cited metric in SaaS go-to-market conversations, and also one of the most routinely misused. Blended CAC, which divides total sales and marketing spend by total new customers acquired, smooths over the differences between channels, segments, and deal types in a way that makes the business look more efficient than it is.

Consider a business that acquires customers through three channels: inbound organic, paid search, and outbound sales. Organic customers cost almost nothing to acquire once the content infrastructure is in place. Paid search customers cost significantly more. Outbound enterprise deals cost more still, but carry higher ACV. Average those together and you get a number that accurately describes none of them.

The practical consequence is that businesses make resource allocation decisions based on a number that masks the actual economics of each motion. They scale paid search because the blended CAC looks acceptable, not realising that paid is dragging up the cost of every customer including the ones that would have come in organically anyway. This is a version of the attribution problem I spent years watching performance marketing teams get wrong: crediting paid channels for conversions that were already in motion before the ad was served.

Earlier in my career I overvalued lower-funnel performance for exactly this reason. The numbers looked clean. The attribution was direct. It took time to understand that much of what performance channels were claiming credit for was demand that already existed. Organic intent, brand recognition, word of mouth, and a product that people genuinely wanted. The paid channel was often the last step, not the cause. In SaaS, this matters enormously because the cost of that last step can be very high relative to its actual contribution to growth.

Segmenting CAC by channel, by customer segment, by deal size, and by cohort gives you a genuinely useful picture. It also tends to produce uncomfortable conversations, which is probably why it does not happen as often as it should. If you are trying to build a more honest view of your go-to-market economics, the Go-To-Market and Growth Strategy hub covers the frameworks that make those conversations productive rather than just painful.

The Retention Cost Problem Nobody Budgets For

SaaS businesses talk constantly about the cost of acquiring customers and almost never with the same rigour about the cost of keeping them. In a subscription model, retention is not a nice-to-have. It is the mechanism by which the business eventually becomes profitable. The LTV:CAC ratio only works if the LTV side of the equation is real, and LTV is only real if customers stay and expand.

Customer success is expensive. Onboarding is expensive. The product investment required to reduce churn is expensive. Most of these costs sit outside the sales budget and outside the cost of sales calculation, which creates a false picture of acquisition efficiency. A business might have a CAC of £3,000 per customer and feel comfortable with that number relative to an ACV of £12,000. But if that customer requires £4,000 of customer success resource in year one to reach adoption and reduce churn risk, the economics look very different.

The businesses that manage this well tend to be the ones that think about cost of revenue rather than cost of sales in isolation. They include the full cost of generating and retaining a pound of ARR, not just the cost of closing the initial deal. This broader view tends to accelerate decisions about product quality, onboarding design, and customer segmentation that a narrow sales-cost view would never surface.

It also changes the conversation about which customers are worth acquiring. A customer segment that closes quickly, onboards easily, and churns rarely looks very different in a full cost-of-revenue model than it does in a simple CAC calculation. Some of the most expensive customers to retain are also the easiest to close, which is a structural problem that only becomes visible when you look at the whole picture. BCG’s work on commercial transformation makes a similar argument about the cost of serving different customer segments and why go-to-market design needs to account for the full revenue lifecycle.

Where Go-To-Market Design Drives Cost Up

Most cost-of-sales problems are go-to-market design problems. The cost is a symptom. The cause is a motion that was built for a different stage of the business, or a customer segment that was never quite right, or a sales process that requires more human intervention than the deal economics can support.

The most common structural issue I have seen is a mismatch between deal size and sales motion. Enterprise sales processes applied to SMB deals. High-touch onboarding for customers whose ACV cannot support it. Sales cycles that run to 90 days for contracts worth £5,000. These mismatches are often inherited from the early days when the business was proving the model and every deal mattered regardless of the cost to close it. They survive because nobody has formally reviewed whether the motion still fits the segment.

The second most common issue is an ICP that has drifted. The ideal customer profile that shaped the original go-to-market design was probably right for an earlier version of the product. As the product evolved, the best-fit customer changed, but the sales motion, the messaging, and the channel mix did not. The result is a pipeline full of prospects that are technically closeable but expensive to close, slow to onboard, and more likely to churn. Vidyard’s analysis of why GTM feels harder points to exactly this kind of compounding misalignment as a primary driver of rising acquisition costs.

The third issue is over-investment in intent capture at the expense of demand creation. This is the performance marketing trap applied to SaaS. Businesses optimise for the bottom of the funnel because the attribution is cleaner and the feedback loops are faster. But the pool of in-market buyers at any given time is finite. Once you have captured most of it, the cost per acquisition starts to rise because you are competing harder for a smaller group of prospects. Creating demand, reaching new audiences, and building the kind of brand recognition that means prospects arrive already warm, is the only sustainable way to grow the top of the funnel without proportionally growing the cost of sales. BCG’s research on go-to-market strategy and pricing reinforces this point in the context of B2B commercial design.

How to Diagnose a Rising Cost of Sales

A rising cost of sales is a signal, not a verdict. The diagnosis matters more than the number itself. Before making any structural changes, it is worth understanding which component is driving the increase and why.

Start with the basics. Is the increase driven by headcount, by marketing spend, by commissions, or by a combination? Has revenue growth slowed while costs have remained constant, which is a different problem from costs actively accelerating? Are specific channels becoming more expensive while others hold? Is the issue concentrated in a particular customer segment or deal type?

The answers to these questions will point to very different solutions. A headcount-driven increase might reflect over-hiring ahead of revenue, or it might reflect a sales motion that requires too many people per deal. A marketing-spend-driven increase might reflect channel saturation, poor targeting, or a measurement problem where spend that was always inefficient is only now visible. A commission-driven increase might reflect a comp plan that was designed for a different deal mix.

One framework I have found useful is to map cost of sales against the sales cycle length and win rate for each segment and channel simultaneously. Businesses that do this often find that their most expensive acquisition channel also has the longest sales cycle and the lowest win rate, which is a straightforward case for reallocation. But they also sometimes find that their highest-cost motion has the highest LTV, which changes the conversation entirely. Forrester’s intelligent growth model makes the case for this kind of multi-dimensional view of commercial efficiency rather than optimising on any single metric.

The diagnosis phase is also where honest conversations about the ICP need to happen. I have sat in enough go-to-market reviews to know that ICP is often treated as a founding document rather than a living hypothesis. The businesses that manage cost of sales well tend to revisit it regularly, with actual data about which customers close fastest, onboard most successfully, expand most reliably, and churn least. That data rarely matches the original ICP perfectly, and the gap between the two is usually where the cost problem lives.

What Good Cost of Sales Management Actually Looks Like

Reducing cost of sales is not primarily a cost-cutting exercise. Businesses that approach it as a cost-cutting exercise tend to cut sales headcount, reduce marketing spend, and then wonder why revenue growth also slows. The goal is to improve the efficiency of the revenue engine, not to make it smaller.

The most effective levers tend to be: tightening the ICP so that sales effort is concentrated on prospects most likely to close and retain; improving the quality of pipeline so that conversion rates rise without requiring more sales activity; reducing sales cycle length through better qualification and more effective enablement; and shifting the channel mix toward lower-cost, higher-intent acquisition sources where the economics are genuinely better.

Product-led growth is worth considering seriously in this context. Where the product can do part of the sales job, the cost of converting a prospect falls significantly. Free trials, freemium tiers, and self-serve onboarding all reduce the human intervention required per deal. The caveat is that product-led growth requires a product that is genuinely good enough to sell itself in the early stages, and an onboarding experience that does not require hand-holding. Neither of those is free to build, but the long-term cost economics tend to be substantially better than a high-touch sales motion at scale.

The other lever that is consistently underused is brand. A business with strong brand recognition in its category generates inbound pipeline at a lower cost than one that has to create awareness from scratch for every prospect. Building that recognition requires investment in demand creation rather than just demand capture, which means being willing to spend in parts of the funnel where the attribution is less clean. That is a harder case to make to a board that is focused on short-term cost efficiency, but it is the right one. The analogy I keep coming back to is the clothes shop: someone who walks in already knowing the brand and wanting to try something on is far more likely to buy than someone intercepted on the street. Getting people to the door is the work that most SaaS businesses underinvest in.

For a broader view of how cost of sales fits into the wider go-to-market picture, the articles across the Go-To-Market and Growth Strategy hub cover the commercial frameworks that connect acquisition efficiency to sustainable revenue growth.

The Measurement Trap

One final point that deserves its own section: the way you measure cost of sales shapes the decisions you make about it. Measurement is not neutral. The choice of what to include, how to attribute, and which timeframe to use will produce different numbers and point toward different interventions.

Businesses that measure cost of sales on a short time horizon tend to optimise for immediate efficiency at the expense of long-term pipeline health. Businesses that exclude customer success costs tend to underestimate the true cost of revenue. Businesses that use blended CAC tend to misallocate budget between channels. None of these are failures of intelligence. They are failures of measurement design.

The goal is not perfect measurement. Perfect measurement does not exist in go-to-market. The goal is honest approximation: a view of cost of sales that is consistent, complete enough to be useful, and designed to surface the decisions that matter rather than to produce a number that looks good in a board deck. Semrush’s analysis of growth strategy examples touches on this tension between metric optimisation and genuine commercial health in ways that are worth reading alongside any internal cost-of-sales review.

I have judged the Effie Awards, which exist to recognise marketing effectiveness rather than just creativity or spend efficiency. The entries that stand out are almost always the ones where the team understood the full commercial picture, not just the channel metrics. Cost of sales, properly understood, is an effectiveness question as much as an efficiency one. The businesses that treat it that way tend to make better decisions.

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