You're Paying for Shelfware: Half Your Software Seats Are Dark

The Cheapest Money You Are Not Saving

Every owner I know is hunting for margin in the hard places. Renegotiating rent. Squeezing a supplier. Deciding whether to backfill a role. Meanwhile there is a pile of money sitting in the software budget that requires no negotiation, no layoff, and no vendor's permission to recover, and almost nobody goes and gets it.

It has a name in the industry: shelfware. Licenses you pay for that no one logs into. When firms actually audit their software, the number that keeps coming back is that somewhere around half of the seats they are paying for are not being used in a given month. Half. That is not a story about one bloated enterprise. It is the base rate.

Here is what's actually happening, and why it is so easy to miss.

Per-Seat Pricing Is Designed to Drift Upward

The dominant way software is sold now is per seat, per month. It sounds fair and it bills quietly, and those two facts are exactly why it leaks.

Think about how a seat gets added. Someone new starts, so you buy them a license for the CRM, the project tool, the design app, the help desk. Now walk through how a seat gets removed. Someone leaves, and, most of the time, nothing happens. The account sits there. The card keeps getting charged $30 or $50 or $80 a month for a person who does not work here anymore. Adding seats has an obvious trigger, a new hire. Removing them has no trigger at all, which means the default behavior of every per-seat contract is to ratchet up and never come down.

Multiply that by a dozen tools and a few years of turnover and you get the standard picture: a subscription base that reflects the largest your team has ever been, not the size it is today. You are not paying for what you use. You are paying for the high-water mark.

Put a Number on Your Own Version of This

Abstractions do not move anyone, so make it concrete with numbers a ten-person shop would recognize.

Say you run five core SaaS tools. The CRM at $50 a seat, a project manager at $15, a design tool at $30, a help desk at $40, and a document-and-signature product at $25. That is $160 per person per month if everyone uses everything, which they never do. Now suppose, conservatively, that a quarter of your provisioned seats are dead, former employees, a tool the sales team abandoned, three logins for a "trial rollout" that never rolled out. On a $3,000 monthly software bill, a quarter is $750 a month. Nine thousand dollars a year, going out the door for access nobody is using.

Nine thousand dollars is a real number. It is a decent chunk of a hire, a marketing budget, a very good year-end bonus. And recovering it does not require selling anything or firing anyone. It requires one afternoon and a willingness to cancel things.

The Vendors Are Not Going to Remind You

Nobody at your CRM company is going to email you to say it looks like four of your twelve seats have not logged in since spring, would you like to drop them. That email costs them money. The entire commercial logic of per-seat SaaS runs on the fact that provisioning is easy and deprovisioning is your problem, and they have built the billing to be exactly frictionless enough that you never feel the individual charge.

This is a cousin of the pricing behavior I wrote about when vendors bolted AI features onto your plan and raised the rate. The mechanism is the same: the increase is quiet, it is defensible line by line, and it relies on you not adding it up. The AI markup shows up on the invoice. The shelfware never even announces itself, because from the vendor's side an unused seat and a used one bill identically.

The Once-a-Quarter Habit That Fixes It

You do not need software to manage your software, and you should be suspicious of anyone selling you a SaaS product to find your wasted SaaS. For a small business the whole cure is a recurring calendar hold, ninety minutes, once a quarter, and a boring checklist.

Pull up each tool's user list and its billing page side by side. For every tool, ask two questions. First, is every name on this list still an employee, because the departed ones are pure waste and should have been cut the day they left. Second, of the people who are still here, who has actually logged in during the last sixty days, because most admin panels will tell you the last active date if you look. The seats attached to nobody get cancelled outright. The seats attached to someone who never signs in get a conversation, do you actually use this, before they get cancelled too. That is the entire method. The reason it works is not that it is clever. It is that nobody does it, so the savings just accumulate quarter over quarter until someone finally looks.

Count the Seats Before You Cut the People

The thing that gets me about shelfware is the order businesses tackle costs in. When money gets tight, the software audit is somehow the last place people look, well after they have started sweating payroll and squeezing the coffee budget. It is backwards. The dead seats are the one cost you can cut with zero human consequence, no morale hit, no lost capability, no awkward conversation, and it is sitting right there in the billing portal you already have the password to.

Go count your seats. Match them against your actual roster and your actual login data. Whatever you find is money that was leaving every single month, and the only reason it kept leaving is that removing a seat, unlike adding one, never had anyone whose job it was to notice.