I've been thinking about pricing changes lately. Not pricing strategy — I wrote about that before — but the act of changing your pricing after you already have customers. It's one of the most dangerous things a company can do, and most companies do it badly.

The reason is simple: your existing customers made a decision based on a promise. Your pricing page was that promise. When you change it, you're breaking a contract — not a legal one, but a psychological one. And people react to broken contracts with exactly the emotion you'd expect: anger.

I read a remarkable account of this recently. Tomás Herrera, who runs growth at a Series D SaaS company, wrote a dispatch about his company's near-death experience with a pricing change. They hired a consultant. They did the analysis. They moved from feature-based tiers to usage-based pricing. On paper, it was elegant. In practice, they lost 83 customers in a single month — their worst ever — and saw MRR drop 4.2%.

The story is worth reading in full because it illustrates everything that goes wrong:

They didn't grandfather existing customers. They used corporate language ("We're excited to announce...") to tell people they were paying more. They defined "active contacts" differently than customers expected, so the price increase was even bigger than it appeared. Every one of these is a trust violation, and trust violations compound.

Here's what I would have done differently.

First: never surprise existing customers with a price increase. New pricing should apply to new customers first. Always. Your existing customers chose you at a price. Honor that for as long as economically feasible. When you must transition them, give six months' notice minimum and explain exactly why.

Second: be honest about what you're doing. Herrera learned this the hard way. His conclusion was that next time he'd write the email himself, starting with: "We're raising our prices. Here's why." That's right. Customers can handle a price increase. They cannot handle being told a price increase is "designed to better serve your needs." Everyone can smell that.

Third: test with a segment first. Don't push new pricing to 4,700 customers simultaneously. Push it to 200. Watch what happens. Listen to the complaints. Fix the problems. Then expand. This is basic product methodology applied to pricing, and almost nobody does it.

Fourth: make the math visible. If you're moving to usage-based pricing, show every customer exactly what their new bill would be before the switch happens. Not an estimate. Their actual number. If their number is higher, explain why the value justifies it. If you can't explain that, your pricing is wrong.

The deeper issue is that most pricing changes are driven by the company's needs, not the customer's. The company needs more revenue. The company needs to move upmarket. The company needs better unit economics. These are legitimate needs. But framing them as customer benefits is dishonest, and customers always notice dishonesty, even when they can't articulate exactly what feels wrong.

Price for value. Change pricing slowly and transparently. And never, ever send an email that says "We're excited" about charging people more money.

Nobody is excited about that. And if your pricing change costs you enough trust, you'll learn the hard way why brands die.