What Does SaaS Sprawl Actually Cost a $5–50M Company?
SaaS sprawl costs a $5–50M company three ways: the subscription line itself (typically somewhere between fifty thousand and half a million dollars a year at this size), the dead weight inside every tool you keep, and the drag of running fourteen dashboards for a business that needs four. The first cost is on your P&L. The other two never show up anywhere — which is exactly why they persist.
One rule before the math: every number below is either drawn from this site's own copy or clearly-flagged illustrative arithmetic you should re-run with your own figures. No invented benchmarks, no "studies show." Your AP report is the only study that matters.
What is SaaS sprawl, exactly?
SaaS sprawl is what happens when a company adds subscription software faster than it removes it — for years. Nobody decides to have sprawl. It accretes: a tool per problem, a tool per department head, a second CRM per acquisition, and per-seat contracts that quietly kept billing after the people in those seats left. At $5–50M you're big enough to have accumulated a serious stack and — usually — still too small to have a procurement team whose job is pruning it. That's the gap this cost lives in.
Cost #1: The line item you can see
Start with the visible number. The founder-architects this audit was built for typically spend $50K to $500K a year on software subscriptions. If you don't know where you fall in that range, that itself is a finding — pull twelve months of AP and card statements and total every recurring software charge. Not the tool list from memory. The list from the money. It nearly always comes back longer and higher than the founder guessed, because sprawl's signature move is hiding on three different credit cards.
Illustrative math — swap in your own figures: a 60-person company carrying 35 recurring tools at an average of $5,000 a year is at $175K annually. Nothing scandalous. That's just what a decade of "it's only $400 a month" adds up to.
Cost #2: The dead weight inside the tools you keep
The premise of this site's scorecard is blunt: for most stacks, the large majority of what you're paying for — features, seats, whole tools — is dead weight. Your team logs into each platform for maybe four features apiece, while the pricing assumes you use forty. You don't have to accept that premise on faith; the audit exists so you can measure your own percentage. It decomposes into three leaks:
- Ghost seats. Per-seat pricing on people who left last quarter, contractors from a project that ended, and licenses that got provisioned "just in case." Illustrative: a $50/seat/month tool with 30 seats paid and 18 active is leaking $7,200 a year on that one line.
- Feature bloat you're billed for. The four-features problem. When a plan tier exists mainly to gate the one feature you actually use, you're funding the vendor's roadmap, not your workflow.
- Redundant tools. Two acquisitions later you have six overlapping CRMs and a finance team that stopped opening the vendor spreadsheet because it was too depressing. Every overlap is a full subscription doing a fraction of a job.
Cost #3: The drag that never hits the P&L
The third cost has no invoice. It's the operating tax of the stack itself: every tool is another login, another place data fragments, another integration to babysit, another dashboard someone has to reconcile against the other dashboards. Context lives in fourteen places, so nobody trusts any single one of them. This is the cost that turns a software problem into an architecture problem — and it's the reason the fix is a redesign of how work gets done, not a coupon at renewal. The shape of that redesign — a short list of load-bearing platforms plus a bench of lean agents — is covered in owned AI systems vs per-seat SaaS.
How do the three costs compare?
| Cost layer | Where it shows up | How you find it |
|---|---|---|
| The subscription line | P&L, spread across cards and AP | Twelve months of statements, one total |
| Dead weight | Nowhere — it's inside "legitimate" spend | Seats paid vs active; features paid vs used |
| Operating drag | Slower decisions, fragmented data, redo work | Count the dashboards a decision touches |
Why doesn't anyone catch this earlier?
Because every individual subscription passes the sniff test. $400 a month is below every approval threshold you have, each tool has one defender who genuinely uses it, and cancellation carries switching-cost fear while renewal is the default that requires zero meetings. Auto-renew does the rest. The result is a ratchet: spend only moves one direction until someone — usually the CFO — finally asks the question out loud. If your CFO is already asking, you're on schedule, and the answer they want is a ranked list, not a shrug. How to audit your SaaS stack before renewal season walks the whole process.
What's the number worth recovering?
Run the shape on your own stack. Say the AP pull comes back at $200K a year. If even a quarter of it is ghost seats, redundant tools, and subscriptions nobody can name a use case for, that's $50K of direct cash back on the P&L — no headcount change, no capability lost, most of it recoverable at the next renewal cycle. The replace bucket comes on top of that: tools with real usage but only four features touched, where a lean agent does those four jobs at a fraction of the seats and cost. Whether your number is half mine or triple it, it's knowable in about six minutes of scoring — which is the entire point of the scorecard.
FAQ
How much do $5–50M companies typically spend on SaaS?
There's no universal figure worth quoting — spend varies wildly with industry and headcount. The businesses this audit is built for typically sit somewhere between fifty thousand and half a million dollars a year. The honest answer for your company is in your own AP report, which takes about an hour to pull and almost always comes back higher than the founder guessed.
Is the "80% dead weight" claim real?
It's the working premise of this site's scorecard, not a lab-verified statistic — and you shouldn't take either on faith. The audit exists so you can measure your own number: score each tool on seats paid versus seats active and features paid versus features used, and see what your actual dead-weight percentage is. For most stacks that have grown for five-plus years without a cull, it's uncomfortable.
Isn't SaaS spend just a cost of doing business?
Some of it is — the load-bearing tools with deep integration or a compliance moat earn their line. The cost-of-doing-business argument fails for the rest: seats for people who left, tools that duplicate each other after an acquisition, and subscriptions nobody on the current team can name a use case for. Those aren't costs of doing business; they're costs of not looking.
What's the fastest way to find out what my stack costs?
Pull twelve months of AP and card statements and total every recurring software charge — not the list from memory, the list from the money. Then run the free twelve-question scorecard on your top tools. It ranks each one kill, replace, or keep, and puts a dollar figure on the savings.