We Changed Our Pricing and Almost Died

San Francisco · January 14, 2026

On November 4th, 2025, at 6:00 a.m. Pacific, we pushed a new pricing page to production. By noon, the support team had received 214 emails. By 5 p.m., the CEO was on the phone with our three largest customers, two of whom were threatening to leave. By midnight, I was sitting in my apartment, looking at a cancellation dashboard that was updating in real time, watching the numbers tick up like a scoreboard nobody wanted to see.

We had changed our pricing. And the market was telling us, in the loudest possible terms, that we had gotten it wrong.

The pricing change wasn't impulsive. That's the thing I want to be clear about. We spent three months on it. We hired a pricing consultant — $28,000 for a six-week engagement — who analyzed our usage data, interviewed twenty customers, benchmarked our competitors, and produced a forty-page deck that recommended what he called "value-based tiering."

The old pricing was simple. Three tiers: Starter at $49/month, Professional at $149/month, Enterprise at $399/month. Each tier had a set of features. You picked the one that matched your needs. It was clean, it was easy to understand, and it had one fundamental problem: 68% of our paying customers were on the $49 tier, and 71% of our revenue came from the 12% of customers on the $399 tier.

We had an inverted pyramid. Lots of customers paying a little, a few customers paying a lot, and a massive gap in the middle that nobody was filling.

The consultant's solution was to move from feature-based tiers to usage-based pricing. Instead of paying for features, customers would pay for "active contacts" — the number of people they were tracking, engaging with, or managing through our platform. The tiers became: up to 1,000 contacts for $79/month, up to 5,000 for $199/month, up to 25,000 for $499/month, and custom pricing above that.

On paper, this was elegant. It aligned price with value — the more you used the product, the more you paid, and the more value you presumably got. The consultant showed us projections: average revenue per account would increase 34%, from $127 to $170. The middle tier would capture the customers who were using more than the $49 tier justified but less than the $399 tier included. ARPU up, churn down (because price tracks value), net revenue retention above 100%.

We believed him. We shouldn't have — not because he was wrong about the theory, but because theory and execution exist in different universes.

Theory and execution exist in different universes, and the space between them is where companies go to die.

The first problem was the email. We sent a pricing change notification to all 4,700 paying customers at 6:00 a.m. The email was written by the marketing team, approved by legal, and reviewed by the CEO. It was professional, clear, and completely tone-deaf.

It opened with: "We're excited to announce a new pricing structure designed to better serve your needs." Customers don't want to hear that you're excited about charging them more. They want to hear why this is good for them, specifically, with numbers they can verify.

The second problem was that we hadn't grandfathered existing customers. Everyone moved to the new pricing on their next billing cycle. The consultant had recommended this — "rip the bandaid off," he said — and we'd agreed because the alternative, running two pricing systems simultaneously, was an engineering nightmare.

This meant that a customer who'd been paying $49/month for two years, who had 3,200 contacts in the system, was about to see their bill jump to $199/month. A 306% increase. Overnight.

The third problem was the contact counting. Our system counted "active contacts" differently than most customers expected. We included contacts in archived campaigns, contacts imported but never used, contacts from integrations that synced automatically. A customer who thought they had 800 contacts actually had 4,300 by our definition, which put them in the $199 tier instead of the $79 tier.

By the morning of the second day, we had 47 cancellation requests in the queue. Our monthly average was 12. The support team was working double shifts. The Slack channel #pricing-fire — hastily created at 10 p.m. the night before — had 892 messages.

I pulled the data. Of the 47 cancellation requests, 31 were from the $49 tier — our largest customer segment. They weren't our highest-revenue customers, but they were our most vocal. Several had tweeted. One had written a blog post titled "How [company name] just alienated their entire user base." It had 4,000 views by noon.

"We have to roll it back," I told the CEO.

"If we roll it back, we look incompetent," she said.

"If we don't roll it back, we lose 200 customers this month."

"What's your estimate on revenue impact?"

"If churn hits the rate I'm projecting — and this is conservative — we lose $38,000 in MRR this month alone. That's $456,000 annualized."

She sat with that number for about ten seconds. "What if we compromise? Grandfather existing customers for six months, new pricing only for new signups."

"That buys us time. It doesn't fix the contact counting problem."

We announced the compromise on November 8th. Existing customers would stay on their current pricing for six months, then transition to the new pricing with a 20% loyalty discount. New signups would see the new pricing immediately. And we redefined "active contacts" to exclude archived and unsynced contacts, which reduced most customers' counts by 30-50%.

The support queue dropped from 89 open tickets to 34 within two days. The cancellation rate slowed. The blog post guy updated his title to "How [company name] almost alienated their entire user base (update: they listened)." Progress, of a sort.

But the damage was done in ways that wouldn't show up in the numbers for months. Trust is like that. You can lose it in a day and spend a year earning it back, and even then it's not the same. Three enterprise prospects who were in late-stage negotiations paused their evaluations. Two cited "pricing stability concerns." One said, simply, "We need to know you're not going to change the deal after we sign."

Trust is like that. You can lose it in a day and spend a year earning it back, and even then it's not the same.

It's been two months since the pricing change. Here's where we are:

We lost 83 customers in November, our worst month ever. December was better — 41 cancellations, which is still above baseline but trending down. January is on pace for 28, which is almost normal.

MRR dipped to $2.87 million in November — a 4.2% decline — and has since recovered to $2.92 million. The new pricing is working for new signups: ARPU for customers who signed up in December is $168, compared to $127 under the old model. But the transition period for existing customers is still six months away, and I have no idea what will happen when those bills actually change.

The consultant, when I called him to debrief, said: "The theory was sound. The execution needed more runway."

He's right. We should have soft-launched the pricing to a segment. We should have tested the contact counting methodology with actual customers before going live. We should have written the announcement email from the perspective of someone who was about to pay more, not someone who was excited about a new pricing model.

Most of all, we should have been honest about what we were doing. We were raising prices. That's not a crime. Companies raise prices. But we wrapped it in the language of "value alignment" and "better serving your needs" and it felt, to customers, like being lied to.

The next time we change pricing — and there will be a next time, because pricing is never done — I'm going to write the email myself. And it's going to start with: "We're raising our prices. Here's why, here's when, and here's what we're doing to make sure you still get more value than you pay for."

Direct. Honest. No consultants.

We'll see if the CEO lets me send it.