A prospect I'd been talking to for weeks sent me a message yesterday that's going to change how I price MVPs for the rest of this year.

He said he'd built half the project himself in under 30 hours using Claude Code. For the rest, he was open to hiring, but only at "pre-AI prices."

That's his exact phrase. Pre-AI prices.

Wrong move one: reprice down to match the new floor. You quote $40K, the market says $10K, you take $10K to keep the lights on. This works for one quarter. Your margins approach zero,your team works the same hours for less money, and you compete head-to-head with offshore agencies that have lower cost bases than you ever will. By month nine you're losing money on the projects you said yes to.

Wrong move two: pretend it isn't happening. Keep quoting $40K to a market that knows AI cut the build cost by 80%. Your close rate halves. Then halves again. By month six your pipeline is dry and you're explaining to your team why nobody's signing.

Wrong move three: panic-pivot to a vertical you don't understand because someone on Twitter said specializing in niche is the answer. They are, eventually. They are not the answer this Tuesday. That pivot takes 12 to 24 months of vertical learning, real client work, and published content. It's the right move if you commit. It's not the move for this month's revenue.

Four moves. Pick one. Start this week.

He's right that Claude Code can ship the happy path of an MVP in 30 hours. He's wrong that this is the whole job. The happy path is roughly 30% of what production software actually is. The other 70% is auth that doesn't leak tokens, payments that don't double-charge, monitoring that catches errors before customers do, deploys that don't take down the database, security that doesn't fail an audit, and the dozen other things that turn working code into a product the founder can run a business on.

Stop selling "build the MVP." Sell "ship a product the founder can actually run." The deliverable is different. The price is different. The buyer who wanted his $10K vibecoded MVP is not your buyer. The buyer who needs that MVP to take real customers and real money is.

If you use Claude Code in your work, say so. Buyers like him know. They have used the tools themselves. Pretending you are not using AI to ship faster is the most insulting move you can make.

Better move: be explicit. Tell the prospect:

"I work with AI-assisted tools. Here's what 40 hours of my work delivers that 40 hours of you on Claude Code alone will not: production hardening, security review, and the failure-mode work most founders skip until they get burned."

Then price that 40 hours at a number that reflects the years of pattern recognition that tells you which failure modes matter.

The reason he could ship the happy path himself is that the happy path is well-documented and AI-friendly. Auth from scratch, payments with proper tax handling, deploys with rollback, monitoring with alerting, and security review for a product taking customer data are not the kind of thing Claude Code ships in 30 hours.

Each of those is a $3K to $8K productized service. Sell them individually or as bundles. This works for two reasons. Each one is small enough that the buyer doesn't need a $40K
commitment to start. Each one is specialized enough that the founder cannot replace you with Claude Code over a weekend.

The founder who tells you they'll vibecode the MVP is still your buyer. They just need a different thing from you now. Six months from now, when their vibecoded MVP has 50 paying customers and is breaking, they need someone who can take it production-grade without rewriting it from scratch. That's a $20K to $50K engagement. They will not call the freelancer who built the happy path for $5K. They will call you, if you stayed in the conversation.

The move is not to walk away when they say they'll vibecode it. The move is to say "good luck, here's what to watch for, ping me when you're ready for the next stage." Then write down their name and follow up in three months.

The next prospect to tell you your prices are pre-AI prices is going to come within the next 30 days. Here is what to actually say.

"You're right that the floor of what an MVP costs has dropped. The number you're seeing reflects what it actually takes to ship something you can put real customers on, not what it takes to get the happy path working. If you want the happy path, Claude Code is genuinely faster and cheaper than I am. If you want a product that doesn't leak tokens, double-charge cards, or fail an audit when you raise your seed round, here is what that work looks like and what it costs."

That answer either closes the deal at your number, or it loses the deal cleanly. Both are better than dropping your price by 60% to chase work you'll regret saying yes to.

Pick one of the four moves. Apply it to the next quote you send out. Hit reply and tell me what you tried and what happened.

Petar