Where Is Margin Quietly Leaking in B2B SaaS?

In most B2B SaaS companies, margin leaks in three predictable places: cloud infrastructure that scales faster than revenue, customer success and support costs that don't shrink as customers mature, and a go-to-market motion where the blended cost of acquisition creeps up while contract sizes stay flat. Value Chain Analysis finds these leaks by mapping every activity that touches a dollar—from product build to renewal—and asking which ones add value customers pay for and which quietly consume margin without it. This article walks through how to run that analysis on a SaaS business and turn it into an execution plan.

Why SaaS Margin Leaks Are Hard to See

SaaS financials hide leaks better than most industries. Gross margins of 70–80% look healthy on a P&L, so leadership stops digging. But those aggregate numbers blend healthy products with unprofitable ones, efficient segments with expensive ones, and self-serve customers with high-touch accounts that cost more to keep than they generate.

The leak is rarely a single line item. It's the cumulative drag of activities that grew organically as the company scaled—a support tier added for one enterprise deal that became standard, a data pipeline that reprocesses everything nightly whether it needs to or not, a sales-assist step layered onto a motion that was supposed to be product-led. None of these show up as a crisis. They show up as gross margin that plateaus a few points below where the model said it would.

Value Chain Analysis, developed by Michael Porter, is built for exactly this: breaking the business into discrete activities so you can see where cost accumulates relative to the value each activity creates.

Running Value Chain Analysis on a B2B SaaS Business

Porter's framework splits activities into primary activities (those that directly create and deliver the product) and support activities (those that enable the primary ones). Here's how that maps to SaaS.

Primary activities to map:

• Product development — engineering, infrastructure, and the R&D that becomes the product itself.
• Hosting and delivery — cloud compute, storage, bandwidth, third-party API costs, data processing.
• Marketing and demand generation — the top of the funnel and everything that fills it.
• Sales and onboarding — closing, implementation, and time-to-first-value.
• Customer success, support, and renewal — the activities that retain and expand accounts.

Support activities to map:

• Engineering platform and tooling (dev productivity, security, compliance).
• Data and analytics infrastructure.
• Finance, RevOps, and G&A.
• HR and talent.

For each activity, ask three questions:

1. What does this cost us, fully loaded? Include people, tools, and allocated infrastructure—not just the obvious line item.
2. Does the customer pay for the value this creates? Some activities are pure value creators. Others are hygiene. Others are pure cost the customer never sees or values.
3. Does this cost scale with revenue, faster than revenue, or independent of it? Costs that grow faster than revenue are your leaks.

What "good" looks like: hosting cost per customer declines as you scale (you're getting infrastructure leverage). Support cost per account declines as the product matures and self-serve deflects tickets. Sales efficiency—new ARR per dollar of S&M—holds or improves by segment. When any of these move the wrong way, you've found a leak.

The most common findings in this analysis: infrastructure over-provisioning for peak loads that rarely happen; a customer segment where CAC payback stretches past acceptable thresholds and never recovers; and a services or support motion that was priced (or given away) below its true cost to serve.

Turning the Analysis Into an Execution Plan

Finding the leak is half the work. The other half is deciding what to do—reprice the unprofitable segment, automate the support tier, re-architect the pipeline, or exit a product line—and sequencing those moves so cash and morale survive them.

This is where a structured platform can compress the timeline. Disclosure: I work on content for Percision, an AI strategic intelligence platform, so treat this as one option among several. Percision runs your business context through Value Chain Analysis alongside its other frameworks and roughly 60+ financial ratios, producing a board-ready view of where activities consume margin and where the leverage is—typically in minutes rather than the weeks a manual teardown takes. It's positioned as a co-pilot, not an autopilot: it surfaces the analysis and scenarios; your leadership team decides.

It's genuinely useful when you need a fast, defensible first pass across a complex value chain, or when you're prepping a board discussion and want a structured artifact to react to. Broadly, AI-assisted analysis has been shown to speed up knowledge work on well-scoped tasks (see the 2023 Harvard/BCG field study on consultants using GPT-4), though those same findings warn accuracy drops on tasks outside the tool's frontier—which is exactly why the human review step matters here.

When you don't need a platform at all: if your leak is already obvious—one product line, one segment, one runaway cloud bill—a finance analyst with a clean spreadsheet and a week will get you there. If the problem is politically sensitive (who owns the unprofitable segment, whose team gets automated), a human consultant or fractional CFO who can navigate the room is often worth more than any model. Use the tool for speed and breadth; use people for judgment and change management.

FAQ

How is Value Chain Analysis different from just looking at my gross margin? Gross margin tells you that you have a problem. Value Chain Analysis tells you where—it decomposes the business into activities so you can trace cost to specific functions and segments rather than staring at a blended percentage.

What's the fastest leak to fix in most SaaS companies? Usually infrastructure and cost-to-serve, because they're operational and don't require repricing customers. Pricing and segment exits create more value but take longer and carry more risk, so they belong later in the sequence.

Can I run this without exposing sensitive financials to a tool? Yes—you can run the framework manually with your finance team, or use a platform for the structure and populate it with anonymized or ranged inputs. Review any tool's data handling before uploading detailed financials.

Want a structured, board-ready pass at where your value chain is leaking margin? You can explore Percision here and decide whether the speed fits your next planning cycle.