The true cost of a SaaS tool is rarely the sticker price

The price on the pricing page is the part of the cost you can see, and for most tools it is not the largest part. Here is a method for calculating what a tool actually costs, worked through on a real category.

The five costs the pricing page omits

1. Unused seats

The most common form of waste, and the easiest to fix. Count logins over the last 90 days, not licences. In our experience the gap is routinely 20–35%, and the seats that get paid for but never touched cluster in predictable places: people who joined for one project, contractors who finished, and the one department that bought licences and then chose a different tool.

effective seat cost = monthly subscription / active users in the last 90 days

A $10/seat/month tool with 12 seats and 8 active users costs $15 per active user, not $10. That premium is real, and it is the number to negotiate against.

2. Unmodelled overage

Usage-based charges are the single largest source of first-year surprise invoices. They are not a bug: vendors set the included allowance to be comfortable for a typical user and profitable for the vendor once you become typical-plus. Storage, API calls, build minutes, seats above a floor, and egress all follow this pattern.

The calculation that matters is not your current usage. It is your usage at three times your current volume, because that is where you will be in two years if the tool is working. If the answer makes the tool indefensible, you have found your real ceiling before you hit it.

3. Migration and switching cost

You are already paying this and not counting it. Exporting data, validating the export, importing, rebuilding integrations, retraining users, running both systems in parallel, and the risk of discovering mid-migration that something does not export cleanly. For a moderate SaaS tool this is routinely 8–40 hours of internal time, which at a blended $75/hour is $600–$3,000 of real cost.

That cost is a genuine argument for staying put. It is also a real argument for a good migration when the alternative is a vendor that has raised prices twice in a year.

4. The integration surface

Tools that sit in the middle of your workflow are more expensive than their licence fee, because every integration is a dependency that can break. When a vendor changes an API, the cost is not their engineering time. It is the team that now maintains a workaround. Budget for the tool being wrong about something, because it will be.

5. The consolidation discount you are not claiming

Vendors discount aggressively for removing a competing tool. This is pure upside if you are already unhappy, and the discount is frequently larger than the negotiated rate you would get by simply asking. If you are considering leaving a tool, tell the vendor you are leaving before you leave.

Worked example: a design collaboration tool

The gap between $540 and $821 is the whole argument for this exercise. It is also why the same arithmetic applied to the two-year, 8-user, $3,000-integration scenario tends to end with a negotiation rather than a cancellation.

The one-page template

Vendor:
Licences purchased:            ___     Active last 90 days:      ___
Monthly subscription:          $___    Real cost per active user: $___
Included allowance:            ___     Current usage:            ___
Usage at 3x (12 months):       ___     Overage at 3x:            $___
Annual commitment:             $___    Discount available:       ___
Migration cost to leave:       $___
Alternative tools at same job: ___

Fill this in for your top ten vendors by annual spend. In most teams this takes about two hours and reliably surfaces more recoverable money than any single negotiation, because it finds the seats and the overage rather than arguing about a price the vendor has already decided.