Every stack audit turns up something they have not opened in ninety days.

Unused SaaS subscriptions are the tools you still pay for but have not opened in ninety days, and they accumulate one reasonable purchase at a time until the yearly total stops making sense. Every stack audit turns up at least one. The fix is naming them, not shaming yourself for buying them.

• Every stack audit I run turns up at least one tool the owner has not opened in ninety days.
• Unused SaaS subscriptions grow through single reasonable decisions, not one reckless one.
• Companies use only about half the software licenses they pay for, according to independent research.
• Nobody sits down and chooses sprawl. It accumulates while you are busy running the business.
• A ninety-day login check is the fastest way to separate working tools from dead weight.
• Cutting an unused subscription is operational hygiene, not an admission of failure.

Here is the sentence I hear every fourth quarter. “My third quarter blew my fourth quarter tech stack, subscriptions and all of that. So for me to stay under for the year, I have to go and cut shit down.” The person saying it is not careless. They are running a real operation. And somewhere in that operation sits a pile of unused SaaS subscriptions that nobody decided to build.

Unused SaaS subscriptions are the software tools a business keeps paying for but has stopped opening, usually defined as no login activity in ninety days, and they accumulate through a series of individually reasonable purchases until the combined total no longer matches the value the stack returns. That is the whole mechanism. No single bad call. Just forty good ones stacked on top of each other.

I do stack audits for a living, so I see the pattern before the owner does. The tool that solved one problem in March. The platform you bought for a project that ended in June. The annual plan that renewed quietly while you were shipping something else. None of it was waste when you bought it. It became waste when it stopped earning its line on the invoice.

Why do unused SaaS subscriptions pile up without anyone noticing?

Unused SaaS subscriptions pile up because every single purchase was a sound decision at the moment you made it, and software bills itself in small recurring amounts that never trigger the same scrutiny as one large payment, so the stack grows past reason without any one choice feeling unreasonable. Thirty dollars here. Forty-nine there. A trial that became a renewal.

The research backs the pattern I see across engagements. A Zylo analysis found companies use only 49 percent of the licenses they provision and waste an average of eighteen million dollars a year on the unused half. That number is for large portfolios. Yours is smaller. The ratio is the same.

Nobody chooses sprawl. It accumulates one reasonable purchase at a time while you are busy running the business.

A separate Productiv report found that on average only 45 percent of company apps get used regularly, and 56 percent were bought outside of any central oversight. That second figure matters. It means more than half the tools in a typical stack showed up because someone solving a real problem reached for the fastest fix. That is not chaos from carelessness. That is operational chaos from momentum.

What does a stack audit actually find?

A stack audit finds three things in nearly every engagement: a tool nobody has opened in ninety days, two tools that do the same job you are paying for twice, and at least one annual renewal that went through without a single person deciding to keep it. The ninety-day tool is the one that stings.

When I pull a client’s subscription list next to their actual login data, the gap is always wider than they guessed. Flexera’s survey of 465 IT professionals found 29 percent of SaaS spend is underutilized or wasted, and their own analyst notes the real figure runs higher than companies estimate. People underestimate their own dead weight. That is human. You remember buying the tool. You forget you stopped using it.

Here is the shape of what turns up in a typical audit:

How do you run a ninety-day stack audit?

You run a ninety-day stack audit by listing every recurring software charge, pulling the last login date for each one, and flagging anything untouched for ninety days as a candidate to cut, then deciding each flagged tool on purpose instead of letting the renewal decide for you. It takes an afternoon, not a project plan.

Start with your card statement, not your memory. Your memory keeps the tools you like. The statement keeps the tools you forgot. Match each charge to a login date, and the dead weight sorts itself.

1. Export three months of card and payment records. Every recurring charge goes on one list.
2. For each tool, find the last time anyone logged in. Most platforms show this in account settings.
3. Flag anything with no activity in ninety days.
4. For each flag, answer one question. If this renewed tomorrow, would I pay for it on purpose?
5. Cancel the ones where the answer is no. Downgrade the overbuilt ones.

Less is best applies here with no exceptions. A smaller stack you actually use beats a complete stack you half-remember. The goal is not the lowest possible spend. The goal is a stack where every line earns its place.

If a tool renewed tomorrow and you would not choose to pay for it on purpose, you already have your answer.

One caution. Do not cut the tool that quietly holds your automations together because you never log into it directly. Some platforms run in the background, moving data between the apps you do open. Check what a tool connects to before you cancel it. A workflow platform like Make.com might show low direct logins while doing heavy lifting every hour. Low login does not always mean low value. That is the one exception to the ninety-day rule.

Why naming the pattern beats shaming yourself for it

Naming the pattern beats shaming yourself because shame makes you avoid the audit, and avoidance is exactly how the stack grew in the first place, while naming it as a normal accumulation you can reverse turns the audit into routine maintenance instead of a confession. You did not fail. You stayed busy.

The current Flexera State of ITAM report found 43 percent of organizations saw SaaS waste rise over the past year, and only 66 percent have real visibility into what they are even paying for. The people with the most sophisticated operations still lose track. The pattern is not a character flaw. It is what happens when a business moves fast and buys the fix in front of it.

The Productiv research found that 56 percent of a typical company’s SaaS apps are shadow IT, meaning they were purchased outside any central approval process. More than half the stack arrived through side doors, which is why no audit ever finds a stack that someone deliberately designed.

Almost every time I open a client’s subscription list, I find a tool they genuinely forgot they were paying for, and their first reaction is embarrassment. I stop that fast. The tool was a good call when they bought it. The only mistake is the autopilot that kept charging for it after the need ended. I ran one audit where the single forgotten annual plan covered the cost of the whole audit three times over. That is the common case, not the rare one. The dead weight is almost always sitting right there in plain sight, hidden only by the fact that nobody had looked at the full list in one place.

How do I know if I have unused SaaS subscriptions?

Pull your last three months of card statements and match each recurring charge to its last login date. Any tool untouched for ninety days is an unused subscription. Most people find at least one they forgot existed. The statement remembers what your memory conveniently drops.

What counts as an unused software subscription?

A subscription is unused when nobody has logged in or acted on it in ninety days. The one exception is background tools that run automations without direct logins. Check what each tool connects to before you call it dead, because low login activity does not always mean low value.

How often should I run a stack audit?

Run a full stack audit once a quarter and a quick card-statement scan monthly. Quarterly catches annual renewals before they fire. The fourth quarter is the natural deadline because that is when the yearly total comes due and the dead weight shows up in one number.

Is it worth canceling a cheap subscription?

Yes, because the problem is the total, not any single line. Ten small subscriptions nobody uses still add up to a real yearly figure. The savings matter less than the clarity. A stack where every line earns its place is easier to run and easier to grow.

What is the most common thing a stack audit finds?

The most common find is a tool unopened in ninety days that renewed on autopilot. Close behind it are two tools doing the same job. Nearly every audit turns up both. The pattern repeats across engagements of every size, which is exactly why it is worth checking.

Will cutting tools break my workflows?

It can if you cut without checking connections first. Map what each tool feeds before you cancel. Background platforms that move data between apps often show low logins while doing critical work. Cancel the ones that connect to nothing and keep the quiet ones that hold things together.

Pick one afternoon this week. Export your statements, match the logins, and flag everything untouched for ninety days. You do not need a tool to do this. You need the full list in one place and one honest question per line.

Then come tell me what you found. What’s the one tool you keep paying for and never open? I’ll go first. Drop it in the comments and name it without shame. Naming the pattern out loud is how it stops repeating.