I Spite-Coded My Own Scheduling Assistant

I use a scheduling tool called Howie. It's good. You CC it on emails, it coordinates calendars, and books meetings. I was paying $35 a month on a month-to-month plan, which felt fair.

Then I wanted to change the name. Instead of my contacts getting emails from "Howie," I wanted it branded to my business. That's it. A string change.

New price: $145 a month. A $110 jump to swap out a name.

Something snapped.

I call it spite coding (after Larry David’s spite store on Curb Your Enthusiasm). A SaaS company pushes you just far enough that you fire up Claude Code and say "I bet I can build this myself."

And I did. In about an hour, I built Remi, a fully functional email-based scheduling assistant. I email him on an alias on my domain, and he handles proposals, follow-ups, and booking. Zoom links, calendar invites, the whole thing. 4,300 lines of Python. 155 tests. A state machine that tracks threads from first contact through booking. Runs on a cron job every four hours with an instant trigger for new emails.

It works. It actually works well.

This is the new math for anyone who builds or manages software. The cost of building a tool from scratch just dropped by an order of magnitude. What used to take a team and six months now takes one person and a couple hours. I just lived it.

Howie isn't unique. This is a pattern you've seen a hundred times: a product launches with a reasonable price, gains traction, takes VC money ($6M from Sequoia and a16z, in Howie's case), and then starts gating basic features behind premium tiers to justify the valuation.

Rename the bot? Premium. Custom domain? Premium. Faster response times? Premium.

This is what I call dead SaaS tactics. Howie is genuinely good at what it does. The pricing strategy is the problem. It treats minor features as major upgrades, and it only works as long as customers have no alternative.

That assumption is breaking. $285 billion evaporated from software stocks in a single week earlier this month. Analysts are calling it the "SaaSpocalypse." Seat-based pricing has dropped from 21% to 15% of SaaS companies in just twelve months. The ground is shifting under every software business that prices based on what customers can't do rather than what they get.

Here's where I'm supposed to tell you to go build everything yourself. I'm not going to do that.

Three days after I finished Remi, I realized I'd built it too narrowly. I designed it for straightforward 1:1 scheduling. Someone emails, Remi proposes five time slots, they pick one. Done.

But what about group scheduling? What if three people need to find a common time, Doodle-style? I hadn't thought of that. What about a personal scheduling page where someone can just grab a slot on my calendar? Didn't think of that either. My rules engine hardcoded a 9-to-5 window. That works until someone on the West Coast wants a 6pm ET call and Remi says no.

I've already added overrides, custom time windows, day-block scheduling, and link detection since the initial build. The scope crept in 48 hours. And that's the honest truth about spite coding: the build is the easy part. The maintenance, the edge cases, the features you didn't anticipate. That's where the real cost lives.

I'd rather pay $35 a month than maintain this myself. That's the honest answer. What I don't want to pay is $145 a month for a feature that costs nothing to deliver.

And that's the new leverage. You don't have to build everything yourself. But you need to know that you could. Because the moment SaaS companies realize their customers can build alternatives in a morning while browsing the web, the dead SaaS playbook stops working.

Here's what I'm watching:

If you're a business owner and a SaaS tool just jacked your price for a trivial feature, you can probably build it yourself with Claude Code or Cursor in a few days. But go in with your eyes open. You're not just building a tool. You're signing up to maintain one.

And if you're running a SaaS company: your customers can see the code now. Price accordingly.