Go and look at your bank statement for last month as part of a subscription and software spend audit. Not the summary in your accounting software. The actual line items.
I will bet there are at least four charges you cannot immediately explain, and two more you can explain but would struggle to justify to anyone.
That is what the exercise is for. It is a boring, one-afternoon task that almost every founder-led business needs and almost none of them do, because it never feels urgent enough to beat the fires.
This post walks through the whole thing. How to pull the real numbers, the categories that always hide waste, the questions that separate a tool you need from a tool you are simply still paying for, and what to do with the money once you find it.
Key Takeaways
- A subscription and software spend audit typically finds 15 to 30 per cent of recurring tool spend going to software nobody has opened in 90 days.
- Pull the data from twelve months of bank and card statements, not from memory. Most founders can name about half the subscriptions they actually pay for.
- Duplicate capability is the highest hidden cost. Three tools that each send email means three bills for one job, and a team guessing which is real.
- Per-seat pricing quietly grows with your team and never shrinks when someone leaves. Check seat counts against actual headcount every quarter.
- Annual plans hide the worst offenders, because the charge lands once and then disappears for twelve months.
- The real cost of a dead tool is not the monthly fee. It is the data trapped inside it and the half-abandoned process built around it.
- Cancel first, consolidate second, automate third. Automating on top of a messy stack just makes the mess run faster.
Sam, the founder I write most of this for, runs a business somewhere between $300k and $5M. His tools were bought one problem at a time. Something broke, he searched, he found a fix, he signed up. Repeat that for six years, and you get a software stack nobody designed, and nobody owns.
Why Software Spend Grows Without Anyone Deciding
Nobody sits down and decides to spend $2,400 a month on software. It happens one $49 decision at a time.
The pattern is always the same. A specific problem shows up on a Tuesday. Somebody starts a free trial. The trial converts. The problem gets solved, or half solved. Six months later, the person who signed up has moved on, the problem has changed shape, and the charge keeps landing on the card every month like a very patient tenant.
This is not carelessness. It is what happens when purchasing decisions are distributed, and nobody is reviewing the total. Each individual decision was rational. The sum of them is not.
There are four specific mechanics that make software spend grow on its own.
Per-seat pricing that only ratchets up. You add a seat when someone joins. Almost nobody removes a seat when someone leaves, because removing seats is admin and it is nobody’s job. I have seen businesses paying for eleven seats on a tool with six actual users, for two years running.
Annual billing that dodges review. A monthly charge gets noticed eventually. An annual charge lands once, gets coded to “software” in Xero, and vanishes for another year. The tools you most need to question are the ones you see least.
Price creep on tools you already have. Vendors raise prices. They email you about it. You are busy. The email gets archived. Your $79 tool is now $129, and you find out during the audit, eight months later.
Trials that nobody cancelled. The team tests three options for one job. One wins. The other two keep billing because cancelling requires logging back in and finding the settings page, and by then everyone has moved on.
Recent surveys put the average small business somewhere between 20 and 40 separate SaaS applications, and research consistently finds a meaningful slice of that spend sitting completely unused. Gartner has tracked the same SaaS sprawl pattern in larger organisations for years. The scale is bigger there. The mechanics are identical.
How to Run a Subscription and Software Spend Audit
Block three hours. Do not do it in fifteen-minute chunks between calls, because the whole value comes from seeing everything at once.
Step one: pull twelve months of statements
Not three months. Twelve. You need a full year to catch the annual charges, and the annual charges are where the ugliest surprises live.
Export every business bank account and every credit card, including the one you think of as personal but which quietly has three business tools on it. Add any PayPal or Stripe billing history, and check the app store subscriptions on company phones.
Do not work from memory, and do not work from a list you already have. The list you already have is the problem. In every audit I have run or watched, the founder’s list captures roughly half the actual subscriptions.
Step two: build one flat list
One spreadsheet. One row per recurring charge. Columns: tool name, what it does in five words, monthly cost, annual cost, who owns it, when it was last used, and what breaks if you cancel it.
That last column does more work than the rest combined. “What breaks if you cancel this” is the question that separates tools you need from tools you merely have.
Write the “what it does” column in plain English, not the vendor’s marketing language. If you cannot describe the job in five words without reading their homepage, that is a finding in itself.
Step three: get honest about last-used
Most tools will tell you if you look. Check the activity log, the last export, the last automation run, the last record created, the last login date on the admin panel.
Where you cannot check, ask the team in one group message: “Who has opened X in the last month?” Silence is the answer.
Anything untouched for 90 days is a cancellation candidate. Anything untouched for 180 days is a cancellation, not a candidate.
Step four: group by job, not by vendor
This is the step that finds the real money. Sort the list by the job the tool does rather than the name on the invoice. Email sending. Scheduling. File storage. Invoicing. Note-taking. Project tracking. Forms. Reporting. Call handling.
Then count how many tools sit in each group. Three tools doing one job is not redundancy for safety. It is three bills, three logins, three data silos, and a team that has to guess which one holds the truth.
What a Subscription and Software Spend Audit Always Finds
Different businesses, same five findings. After you have run a few of these, the results stop surprising you.
The ghost tools. Software from a project that ended, a staff member who left, or a strategy you abandoned. The trade-specific quoting tool you trialled for two months. The social scheduler from when you were going to post daily. These are pure cancellations with no consequence, and they usually account for the first chunk of savings.
The duplicate stack. You are paying to send email through your CRM, through a standalone marketing tool, and through a transactional service one of your automations quietly depends on. Two of those three could go this week. The same story repeats with file storage, forms, scheduling, and note-taking.
The over-provisioned seats. Every tool with per-user pricing, checked against your actual headcount. This is normally the fastest win, because it requires no cancellation and no process change. You reduce the seat count, and the bill drops next cycle.
The wrong tier. You are still on the plan you needed when you signed up. Sometimes that means paying a premium tier for a feature you used twice. Sometimes it means the opposite, and you are choking a tool you rely on daily by keeping it on a starter plan. Both are worth fixing, and one of them will save money while the other makes you money.
The tool that replaced nothing. This is the expensive one. You bought software to solve a problem, and the old manual process is still running alongside it. Now you pay for the tool and still do the work by hand. That is not a software problem; it is an adoption problem, and cancelling the tool is often the honest call.
Add those up and a typical founder-led business finds somewhere between 15 and 30 per cent of recurring software spend is dead or duplicated. On a $2,000 monthly stack, that is $300 to $600 a month, or $3,600 to $7,200 a year, sitting in plain sight on a statement nobody reads.
The Trap: Cutting Costs Without Fixing the Cause
Here is where most audits go wrong. The founder finds the waste, cancels eight subscriptions, feels good about it, and eighteen months later the exact same pile has rebuilt itself.
The audit is a treatment. It is not a cure. The cause is that nobody owns the stack, nobody reviews the total, and every purchase decision gets made in isolation by whoever hit the problem first.
Three things fix the cause, and none of them is complicated.
One owner. Somebody, ideally not you, holds the list. Every new subscription goes through them. Not for approval theatre, just so the list stays real.
One card. All software on one card or one account. It makes the next audit take one hour instead of three, and it makes the monthly total visible whether you go looking or not.
One calendar reminder. Quarterly. Fifteen minutes. Open the list, check seat counts, check last-used dates, cancel whatever died since last time. Quarterly beats annual because the drift never gets big enough to hurt.
If you want to be properly rigorous, add one rule for new purchases: any new tool has to name the tool it replaces or the manual process it retires. If it replaces nothing, it is a permanent addition to your cost base with no offset, and that needs a much better reason than “it looked good on the demo.”
What to Do With What You Find
The money you recover is real, but it is not the biggest prize.
By the end of a proper audit, you have something you did not have before: a written map of every system in the business, what job each one does, who touches it, and how data moves between them. Most founders have never had that on one page. Building it is uncomfortable, because it shows you exactly how much of your operation is held together by one person remembering how it all connects.
That map is the same thing you need before you automate anything. This is why I run a spend audit before touching automation with a client. Automating on top of a messy stack means building the same process three times across three overlapping tools, then maintaining all three forever. Clean up first, connect what survives second.
Once the stack is lean, the question changes shape. It stops being “which tool should I buy?” and becomes “what work is still happening manually that should not be?”
That is the shift from buying software to building a system: an AI brain that holds your business context, and an AI workforce that does the recurring work. Tools sit inside that. They do not replace it. Most of the businesses I work with end up with a smaller software bill and more output, because the problem was never the number of tools. It was that nobody had designed how they fit together.
For the wider pattern, admin overload in small business covers where the hours actually go, and the 90-day AI automation checkpoint is a useful frame for reviewing anything you build after the cleanup.
Conclusion
A subscription and software spend audit is not glamorous work. It is a spreadsheet, a stack of statements, and three hours you would rather spend on almost anything else. It is also one of the highest-return afternoons available to a founder-led business, because the money is already yours and you are simply choosing to stop giving it away.
Do the audit. Cancel the dead weight. Put a quarterly reminder in the calendar so it never gets this bad again.
Then ask the harder question the audit surfaces. Why is so much of this business held together by tools nobody designed and knowledge nobody wrote down? That is the actual constraint. The software bill is just where it shows up on a statement.
If you want a second pair of eyes on your stack and on what should happen after the cleanup, book a 30-minute Discovery Call. No pitch deck, just a look at what you are running and where the obvious wins sit.
Frequently Asked Questions
How often should I audit my software subscriptions?
Do one full audit working from twelve months of statements, then review quarterly for fifteen minutes. The full audit is the heavy lift because you are building the list from scratch. After that, quarterly checks catch new subscriptions, seat count drift, and price increases before they compound. Annual reviews are too infrequent, because a year is long enough for three or four dead tools to accumulate unnoticed.
How much do small businesses waste on unused software?
Industry research consistently puts unused or duplicated SaaS spend at roughly 15 to 30 per cent of the total software bill for small and mid-sized businesses. In practice, that means a business spending $2,000 a month on tools is usually giving away $300 to $600 of it. The waste is rarely one large subscription. It is eight small ones that individually look too cheap to bother cancelling.
What is the difference between a software audit and a SaaS spend audit?
They overlap, and most people use the terms interchangeably. A software audit sometimes means a licensing compliance check, confirming you hold the right number of licences for the software you run. A SaaS or subscription spend audit is about cost: what you pay for, what you actually use, and what you can cut. For a founder-led business, the spend version is almost always the one that matters.
Should I cancel a subscription if we only use it occasionally?
Weigh the annual cost against how often it gets used and what happens if it disappears. A tool used twice a year for something genuinely important is worth keeping. A tool used twice a year out of habit is not. Where a monthly plan exists, downgrading and re-subscribing when you need it usually costs far less than a year of dormant access.
How do I find subscriptions I have forgotten about?
Bank and credit card statements are the only reliable source. Export twelve months, sort by amount, and work through every recurring charge. Also check personal cards used for business, app store subscriptions on company phones, and PayPal or Stripe billing histories. Do not rely on your inbox, because vendor receipt emails get filtered and buried within weeks of signing up.
Will cutting software spend hurt my team’s productivity?
Only if you cancel the wrong things, which is why the “what breaks if you cancel this” column matters. The tools that hurt to lose are the ones people use daily and will tell you about immediately. The ones nobody notices are the ones that should have gone a year ago. Give the team a week of notice before cancelling anything ambiguous, and you will find out fast.
What should I do with the money the audit frees up?
Put it toward fixing whatever the audit exposed. Most audits surface a process held together by manual work and one person’s memory, and that deserves more attention than the savings do. Recovered software spend is often enough to fund the first real piece of automation, which is a better return than banking the money and letting the stack drift again.
About Octavius
Titus Mulquiney is the founder of Octavius AI, where he builds AI brains and AI workforces for founder-led businesses stuck running everything out of their own head. Twenty years in marketing, ex-Sony product manager, ex-GM Zeal NZ. Based in Auckland, working with operators across NZ, Australia, and the US. Connect on LinkedIn.