If it feels like your software subscriptions rise faster than everything else you pay for, they do. Last year, average SaaS prices rose 11.4% while US inflation ran 2.7% — roughly four times the general rate, and part of a multi-year pattern rather than a one-off. By the same analysis, 72% of SaaS vendors’ revenue growth in 2025 came from raising prices on existing customers, not from winning new ones. One 2026 benchmark puts it concretely: a 50-person company that spent $84,000 a year on software in 2024 is looking at over $112,000 in 2026.
None of this is a scandal. It is a business model behaving exactly as designed, and the design is worth understanding.
A software vendor’s pricing power is your switching cost. Once your data, your history, and your team’s habits live inside a tool, leaving it means months of disruption and real money. The vendor can estimate that cost at least as precisely as you can, and renewal pricing is set against it: the increase will reliably be painful, and reliably be smaller than the cost of leaving. That is not malice. It is arithmetic.
Per-seat pricing adds a second, quieter tax: you pay for headcount, not for value. Hire five people and you buy five more seats of everything, whether or not those five touch the parts of the tool that matter. The software did not get better; the bill got bigger.
And there is a third cost with no line item at all — the fit. An off-the-shelf tool is built for the median customer of its category, and your business is not the median customer of anything. Some part of your operation was reshaped to fit the tool on the day you bought it, and the workarounds have accumulated since. The rent goes up every year; the compromise stays exactly where it was.
To be clear, renting is often still the right call. For commodity processes — accounting, email, payroll — you are the median customer, and the standard tool at the standard price is the correct answer. Anyone who suggests building your own email is selling something.
The calculation is different for the process that makes your company your company: the quoting, scheduling, or delivery that you do differently, and win work because of — the process where the bought tool never quite fit. For two decades, tolerating that fit was still rational, because custom software cost more than the compromise was worth. That price has fallen a long way. The rent, meanwhile, compounds.
So the useful question at renewal time is not “can we absorb the increase.” It is “what exactly are we renting, and would we still choose it today.” For most of the stack the answer is yes — renew and move on. It only takes one no to matter.