Nobody sets out to buy the same software twice. It happens the way most expensive things happen in a growing business: one reasonable decision at a time, none of them wrong on the day.
Marketing needs a design tool, so they buy one on a card. Operations trials a project tracker and it sticks. Someone signs up for a transcription service for one meeting. Three years later there are forty-odd subscriptions, nobody has a complete list, and a meaningful share of the monthly spend is buying nothing at all.
The four kinds of waste
Seats belonging to nobody
The most common and the easiest to fix. Staff leave, IT disables the accounts it knows about, and the tools bought outside that process keep billing. If your offboarding checklist doesn’t include a list of every SaaS product in the business, it can’t possibly be complete.
Overlap
Two products that do substantially the same job, usually adopted by different teams at different times. The subscription cost is the visible part. The hidden part is that your information is now split, so searching is unreliable and onboarding takes longer.
Worth checking specifically: chat, file storage, project tracking, e-signature, and video calls. Those five are where duplication concentrates, largely because every larger platform bundles a version of each.
Tier drift
You bought the higher tier for one feature, or because it was needed at a moment of growth, and never re-examined it. Vendors are not going to write and suggest you downgrade. Check what each tier actually gives you against what you use.
The quiet auto-renewal
Annual subscriptions that renew on a date nobody has in a calendar, for a product whose champion left the business eighteen months ago. These can run for years. I have found subscriptions still billing for tools that were decommissioned before the current finance manager started.
How to run the audit
This takes about half a day and it’s the highest-return half-day available to most businesses.
- Pull twelve months of card and bank statements and extract every recurring payment to a software vendor. Twelve months, not three, because annual renewals hide.
- Export sign-in activity from your identity provider. Anything with no logins in ninety days is a candidate for cancellation or, if it’s business-critical and still unused, a much more interesting conversation.
- For each product, write down three things: who owns it, how many seats are paid for, and how many are genuinely active. The gap between the last two is usually where the money is.
- Group by function. Put every product under a heading like “file storage” or “communication”. Duplication becomes obvious the moment things sit next to each other.
- Diarise every renewal date. Not to cancel, but so the decision happens deliberately rather than by default.
The part that’s actually hard
Finding the waste is straightforward. Removing it is political.
Every duplicate tool has someone who chose it, likes it, and will experience its removal as a judgement on them. If you approach consolidation as a cost-cutting exercise imposed from above, you’ll get resistance, and often quiet non-compliance where the tool keeps being used on a personal card.
What works better: agree the standard for each function as a group, give people a genuine say in which one wins, and set a date. Then actually turn the other one off, because a consolidation that leaves both running has achieved nothing except a longer list.
A word on shadow IT
Everything above treats unapproved tools as a cost problem. They’re also a security and privacy problem, because company data is sitting in a service nobody assessed, under terms nobody read, possibly outside Australia.
The instinct is to ban it. That mostly drives it further underground. The better read is that shadow IT is a signal: people bought something because the approved option didn’t do the job or was too hard to get. Fix that and most of it stops appearing.
What good looks like afterwards
One list of every software product, with an owner, a seat count and a renewal date, reviewed twice a year. That’s the whole discipline. It isn’t sophisticated and almost nobody does it.
The saving is real, but the better outcome is that new purchases start going through a decision rather than a card, which stops the whole thing rebuilding itself over the next three years.