Why per-seat pricing won when it makes no sense

At my last job we sold a data pipeline tool. One customer—a mid-market fintech—had 200 employees but maybe 12 people actually used the product regularly. Still paid for 200 seats. Another customer had 15 people but they hammered the system so hard we probably should've charged them triple. We charged them for 15 seats. I think per-seat won because it's the path of least resistance for *selling*, not because it matches how software actually gets used. You can explain it in thirty seconds to a procurement person. It's transparent on the contract. It scales predictably with headcount, which finance teams already track. And critically, the person writing the check—the CFO or procurement lead—can justify it to their boss. "We have 200 people, we pay for 200 seats" is a conversation that ends. Usage-based pricing requires you to defend *why* someone's bill jumped. The real cost to vendors is that this breaks down constantly at scale. We spent cycles on seat audits. Customers got angry when they hit growth and their bill jumped. We had customers deliberately under-reporting headcount or parking unused licenses. And in the meantime you're leaving money on the table from the power users—the team that could actually afford to pay more because they're the only reason the product gets budget approval in the first place. Usage-based and hybrid models work better when you build the pricing *and the product* for it from the start. Slack's model works partly because slack activity is something you can actually measure and defend. But most SaaS companies didn't do that. They bolted seat pricing onto the go-to-market playbook, it stuck, and now they're sort of stuck with it. Not because it's optimal. Just because changing it mid-flight terrifies finance and sales.