Gross margin is one of the first numbers a VC will pull up against benchmark data, and it’s often the first place a diligence conversation gets uncomfortable.
A founder walks in reporting 78% gross margin, confident that number reflects a healthy SaaS business, and the investor’s model shows comparable companies at their stage sitting closer to 65%. That gap isn’t usually a sign the business is genuinely more efficient than its peers. More often, it’s a sign that what’s sitting inside COGS hasn’t been rigorously defined, and the founder didn’t know there was a question to ask in the first place.
This is one of the most common blind spots Rooled encounters with new clients, and it’s rarely the result of anyone trying to make the numbers look better than they are. It’s simply that gross margin gets calculated once, early on, using whatever categorization felt intuitive at the time, and nobody revisits the definition as the business grows and the cost structure gets more complex. The gap that creates is often ten to twenty percentage points, which is more than enough to change how an investor reads the rest of the story.
The Gross Margin Credibility Problem
The credibility problem shows up the moment an investor compares your reported gross margin against the benchmark data they already have for companies at your stage and business model. If your number is meaningfully higher than the comparable set, the first assumption isn’t that you’ve built something more efficient. It’s that something is misclassified, and that assumption puts the burden on you to explain the gap rather than letting the number speak for itself.
At the root of this is a definitional problem that sounds simple and rarely is in practice: what belongs in cost of goods sold, and what belongs in operating expenses below the gross margin line. There’s no universal answer, and reasonable finance teams can draw this line in slightly different places depending on the business. What matters is that the line gets drawn deliberately and consistently, rather than inherited from whatever felt right when the chart of accounts was first set up, often before anyone on the team had built a SaaS income statement before.
What Should (and Shouldn’t) Be in COGS
For a SaaS business, COGS should generally include hosting and infrastructure costs, customer support directly tied to serving existing customers, onboarding costs for new customers, and any third-party licenses or data costs required to deliver the product. These are the costs that scale with usage and delivery, which is the defining characteristic of a cost of goods sold line versus an operating expense.
For a services or consulting business, COGS should capture labor directly attributable to client delivery, meaning the people actually doing the work customers are paying for, as distinct from sales, marketing, and general administrative labor that supports the business but doesn’t touch delivery directly.
The single most common mistake Rooled sees is excluding customer success from COGS entirely, treating the team as a sales and marketing or G&A expense instead. If a customer success team’s core function is supporting existing customers so they renew and expand, that cost is directly tied to serving the customer relationship, which is exactly what COGS is meant to capture. Moving customer success below the gross margin line, into operating expenses, is one of the fastest ways a startup’s gross margin ends up overstated relative to how investors actually expect the metric to be built.
Why Gross Margin Presentation Varies
Part of what makes this confusing is that there isn’t just one version of gross margin, and different versions are appropriate for different conversations. GAAP gross margin follows strict accounting definitions and is the version your auditors and your financial statements will reflect. Contribution margin strips out a narrower set of variable costs and is often more useful internally for understanding unit-level profitability on a specific product or customer segment. An “adjusted” gross margin, which some companies present to smooth out one-time costs or unusual items, can be a legitimate lens when used transparently, but it’s also the version most likely to draw scrutiny if the adjustments aren’t clearly disclosed.
The right approach is to know which version you’re presenting and why, and to be explicit about it rather than letting “gross margin” mean three different things in three different conversations. An investor who discovers that your board deck margin and your GAAP margin diverge by fifteen points, without a clear explanation of what’s been adjusted and why, will spend the rest of diligence questioning every other number in the deck. The risk of overclaiming isn’t just that one number looks worse once corrected. It’s that it puts every other metric you’ve presented under a more skeptical lens.
Improving Gross Margin: Tactical and Structural Levers
Once the definition is clean, the actual work of improving gross margin splits into tactical and structural levers, and it’s worth being clear about which kind of lever you’re pulling.
Infrastructure efficiency is usually the fastest tactical win, whether that means optimizing cloud spend that’s grown inefficient as usage patterns shifted, or renegotiating vendor contracts that were signed at an earlier, smaller-scale price point. These fixes tend to be available without touching the product or the customer relationship at all.
Pricing architecture is a deeper lever, and it starts with an honest question: are you pricing to cover your true cost to serve, including the support and onboarding costs that belong in COGS? A pricing model built before those costs were well understood can quietly bake in margin compression that no amount of infrastructure optimization will fix on its own.
Customer mix is the most strategic lever of the three, and it requires cohort-level margin analysis rather than a single company-wide number. Some customer segments are dramatically more expensive to serve than others, whether because of support intensity, usage patterns, or contract structure, and a blended gross margin can hide a segment that’s actively dragging the average down. This is exactly the kind of analysis a platform like Aleph makes far more practical to maintain on an ongoing basis, since tracking margin by cohort by hand becomes unwieldy the moment your customer base has any real diversity to it.
Modeling Gross Margin Improvement
The strongest position to be in isn’t a static gross margin number, clean as it might be. It’s a model where gross margin is an output of your other operating assumptions rather than an assumption typed directly into a spreadsheet cell. When margin is modeled as the result of infrastructure cost per customer, support cost per account, and pricing by segment, you can show a credible path from where margin sits today to where it needs to be, with each point of improvement tied to a specific, named driver.
That credible path is what actually earns investor confidence, far more than the current number on its own. A founder who can say “our margin improves from 68% to 76% over the next six quarters as infrastructure cost per customer drops with scale and our pricing catches up to our actual cost to serve” is telling a story an investor can underwrite. A founder who can only say “we expect margin to improve” is asking for trust without giving the underlying reasoning to earn it.