How much can you realistically save by auditing your SaaS stack?

For a team that has never run a structured audit, 20–30% of total SaaS spend is a realistic cut without losing a single tool people actively use. It decomposes roughly as: ~10% from deprovisioning orphaned or ghost seats, ~8–10% from right-sizing over-provisioned plan tiers, ~5–8% from eliminating functionally overlapping tools, and ~5% from converting genuine keepers to annual billing. Stacks that have grown bottom-up with no central owner are often at the higher end. The savings shrink on each subsequent audit — the first one captures the accumulated bloat.

Should I switch my SaaS subscriptions to annual billing to save money?

Only for tools you are confident you'll still use in twelve months. Annual billing typically saves 15–20% versus monthly, which is real money on your core stack. But prepaying a year on a tool you abandon in month four is a negative saving — you've locked in spend you can't recover. The rule: annualize the keepers, keep anything experimental, seasonal, or under evaluation on monthly until it proves itself. Run each tool through a simple 'will we still use this in a year' test before committing.

What are ghost seats and how do I find them?

Ghost seats are paid licenses still being billed for people who have left, changed roles, or never used the tool — the single largest silent leak in most stacks, because per-seat tools keep charging whether or not anyone logs in. Find them by exporting each tool's user list and reconciling it against your current headcount or an active-directory roster, then checking last-login or last-active dates. Anyone inactive for 60–90 days is a candidate for removal. Deprovisioning at offboarding, not at audit time, is the permanent fix.

How often should a company audit its SaaS subscriptions?

A full stack audit once or twice a year, plus a lightweight per-renewal check that's far more effective: review each subscription 30 days before its renewal date, when cancellation is still possible and you have leverage. The recurring discipline beats the occasional purge, because SaaS waste re-accumulates continuously as people add tools and headcount changes. Assigning an owner to a renewal ledger turns a one-time cleanup into a standing control.

How to Audit Your SaaS Stack and Cut 30% (2026)

Q: What is a good SaaS spend per employee?

Spend per active seat is the better diagnostic than total spend because it normalizes across company size. As a rough operator benchmark: under about $50 per seat per month is lean and well-controlled, $50–120 is moderate with room to consolidate, $120–250 is bloated and worth a tool-by-tool review, and over $250 per seat signals an immediate audit. These vary by function — an engineering-heavy org carries more per-seat tooling than a sales team — so track the trend over time more than the absolute number.

The "cut 30% of your SaaS spend" headline is real, not clickbait — but the usual method behind it is too crude to deliver it without collateral damage. "Find the tools nobody uses and cancel them" catches the obvious dead weight and stops there, leaving the structural waste untouched. The cuts that actually add up to a third of your bill come from places a casual scan never looks: licenses assigned to people who left, plan tiers three steps above what you use, and three tools quietly doing the same job. None of those show up as "unused." All of them are pure margin. Let's find them properly.

1. Stop looking at total spend. Per-seat is the diagnostic.

Total SaaS spend is a vanity metric — it only tells you how big you are, not whether you're wasteful. A 50-person company spending $40,000/month might be lean; a 10-person company spending $18,000 is bleeding. The number that exposes the truth is spend per active seat, because it normalizes across company size and turns "is this a lot?" into a question you can actually answer.

As a working benchmark: under roughly $50 per seat per month is lean and well-controlled; $50–120 is moderate, with room to consolidate; $120–250 is bloated and worth a tool-by-tool review; and over $250 per seat is critical — a stack that has grown without anyone steering it. The absolute figure varies by function (engineering carries more per-seat tooling than sales), so the trend matters more than the snapshot. A per-seat number that creeps upward quarter over quarter is the real signal: it means tools are being added faster than headcount, which is almost always decentralized buying with no owner.

Get your per-seat score in two minutes — list your tools and seats in the SaaS Subscription Audit and it grades your stack lean → critical and surfaces the annual billing you're leaving on the table.

2. The three wastes that actually compound

"Unused subscriptions" is one kind of waste, and the least interesting. Three others compound silently and account for most of the recoverable spend:

Ghost seats (orphaned licenses). The single biggest leak. Per-seat tools keep billing for people who left, changed teams, or never logged in — because deprovisioning rarely happens at offboarding. A 10-seat plan where 3 seats belong to former employees is a 30% overpay on that one line, invisible until you reconcile the user list against current headcount.
Tier inflation.Vendors design pricing so that one feature you need sits one tier above the plan that covers everything else you use. You end up on Enterprise for a single SSO requirement or a seat threshold, paying for a dozen capabilities you'll never touch. Auditing which tier-defining feature you actually use frequently reveals you can drop a level.
Functional overlap. Because SaaS adoption is bottom-up, teams accrete redundant tools: Notion and Asana and Linear all managing projects; Zoom, Google Meet, and Slack huddles all doing video; two analytics tools, three places to store files. Nobody chose the redundancy — it grew. Mapping tools to jobs rather than to vendors exposes where one could replace three.

3. Where the 30% actually comes from

The honest version of the headline isn't a single dramatic cancellation — it's a stack of unglamorous recoveries that add up. For a team auditing for the first time, the decomposition usually looks like this:

~10% — deprovisioning ghost seats.The fastest, lowest-risk money. Nobody loses a tool they use; you just stop paying for people who aren't there.
~8–10% — right-sizing tiers. Dropping plans that are over-provisioned for your actual usage, or negotiating at renewal with the usage data in hand.
~5–8% — killing overlap. Picking one tool per job and migrating off the duplicates. Slower, because it touches workflows — but durable.
~5% — annualizing the keepers.Converting tools you're certain about from monthly to annual billing for the standard 15–20% discount.

Stacked, that's a realistic 25–30%without removing a single tool anyone relies on. The point of the decomposition is that you don't need a heroic decision — you need four ordinary ones executed in sequence.

4. The annual-vs-monthly arbitrage — and its trap

Annual billing saves 15–20% over monthly, and the reflexive advice is "switch everything to annual." That advice is half right and quietly expensive. The saving is real on your corestack — the tools you'd bet money you'll still run next year. But prepaying twelve months on a tool you abandon in month four is a negativesaving: you've converted a cancellable monthly cost into sunk, unrecoverable spend.

So the rule is conditional, not blanket: annualize the keepers, keep the experiments monthly.Run each tool through a one-line test — "will we still be using this in twelve months?" If the honest answer is yes, take the annual discount. If it's "probably" or "we're still evaluating," the flexibility of monthly is worth more than the discount. Most stacks have a clear core to lock in and a churning edge to leave liquid.

The SaaS Subscription Audit flags exactly how much each line would save on annual billing, so you can annualize the keepers without doing the math by hand.

5. The renewal ledger — the leverage moment nobody owns

Here is why SaaS waste re-accumulates no matter how thoroughly you purge it once: subscriptions auto-renew, and the only moment you have real leverage — to cancel, downgrade, or renegotiate — is the narrow window before the renewal date. In most companies, nobody owns that calendar. The renewal sails through silently, and the spend persists for another full term.

The fix is structural, not heroic: a renewal ledger— a simple table of every subscription with its renewal date, owner, seat count, monthly cost, and a "still needed?" column, reviewed 30 days before each renewal. That review is where the audit stops being an annual event and becomes a standing control. It's also your negotiating window: vendors discount far more readily 30 days out from a renewal you're visibly willing to walk from than they do mid-term.

6. Two audit lenses: seat-based vs usage-based

A subtlety that trips up otherwise-careful audits: not all SaaS bills the same way, and the two models need different scrutiny.

Seat-based tools (Slack, Figma, Notion, most CRMs) — audit the seats. Ghost seats and tier inflation are the targets. The bill scales with headcount, so the user list is your battlefield.
Usage-based tools (Datadog, Twilio, cloud infra, LLM APIs) — audit the consumptionand the commitment tier. There's no "unused seat" to find; the overspend hides in unthrottled usage, over-committed reserved tiers, or inefficient configuration. A "cancel what you don't use" frame misses this entirely, because you are using it — just more expensively than you need to.

If your AI and infrastructure lines are a meaningful share of spend, they deserve their own consumption audit rather than being lumped into the seat review. The LLM API cost calculator is built for that side of the ledger.

7. The consolidation trap: compute the payback period

Eliminating overlap is the most satisfying cut and the easiest to get wrong. Collapsing three tools into one platform looks like obvious savings — until you account for the migration: data export and import, workflow rebuilding, retraining the team, lost features you only discover you depended on, and the productivity dip during the switch. Those costs are real even though they don't appear on an invoice.

Before consolidating, do the unglamorous arithmetic: how many months of the saved subscription does the migration cost consume? If killing a $200/month overlap takes a two-week migration and a month of reduced output, the payback period might be most of a year — and if you'd have re-evaluated the surviving tool by then anyway, the cut wasn't worth it. Consolidate where the payback is short and the surviving tool is one you're committed to. Leave overlaps that are cheap to keep and expensive to merge.

The playbook

Compute spend per active seat first. It tells you whether you have a problem and how big, before you touch a single subscription.
Reconcile every user list against current headcount. Kill ghost seats. This is the fastest, safest ~10%.
For each tool, name the feature that sets your tier. If you can't, you're probably one tier too high.
Map tools to jobs, not vendors. Where one job has three tools, plan a consolidation — but check the payback period before pulling the trigger.
Annualize the keepers, keep experiments monthly. Take the 15–20% discount only where you're sure.
Build a renewal ledger and assign an owner. Review each line 30 days before renewal. This is what stops the waste from growing back.

Start with step one. The SaaS Subscription Audit scores your per-seat spend and finds your annual-billing savings in one pass. For the payments side of your stack, the Stripe True Fee Calculator does the same for processing costs.

FAQ

Isn't cutting tools just going to slow the team down?

A bad audit does; a good one doesn't. The first three quarters of a typical 30% cut — ghost seats, tier right-sizing, annual conversion — remove zero capability from anyone. Only consolidation touches workflows, and that's exactly where you apply the payback-period test rather than cutting reflexively. Done right, an audit removes spend, not tools people rely on.

What's the difference between a SaaS audit and just cancelling things?

Cancelling is a one-time reaction to obvious dead weight. An audit is a method: measure per-seat spend, classify waste into ghost seats, tier inflation, and overlap, decide each line on evidence, and install a renewal ledger so the result holds. The cancellation captures maybe a third of what's recoverable; the structural work captures the rest and keeps it captured.

Who should own SaaS spend in a small company?

One named person — usually whoever owns finance or operations — even if buying stays decentralized. The owner doesn't approve every tool; they hold the renewal ledger, run the per-seat number, and own the 30-days-before-renewal review. Spend creeps precisely when the answer to "who owns this?" is "nobody."

Are these per-seat benchmarks the same for every business?

No — they're a starting reference, not a universal target. Engineering- and data-heavy organizations legitimately run higher per-seat tooling than sales- or ops-led ones. Use the benchmarks to locate yourself roughly and, more importantly, to watch your own trend over time. A rising per-seat figure is the signal, whatever your baseline.

Pricing benchmarks current as of May 2026. General operational guidance, not financial advice — confirm current plan pricing on each vendor's page before making changes.

How to audit your SaaS stack and cut 30%