So I have this problem where my brain will spend six hours building a Python script to solve a problem that would take three minutes to just look up.
Anyway, I automated my fishing trips.
The Script Nobody Asked For
The system checks barometric pressure trends from 9 PM to midnight, moon phase, wind speed, cloud cover, and rain probability. If the composite score is below 65, it tells me to stay home. It has a kayak safety veto: wind above 15 mph, automatic no. Probability of lightning within 10 miles, hard stop.
I built that. For fishing. From a kayak.
I know what you’re thinking. “Dude, just check the weather.” Yeah. I could. But my brain doesn’t work that way. It works in systems. And honestly? The weather app gives you a number. My script gives you a decision. Those are different things.
(Also the fish don’t care about your weather app. I don’t make the rules.)
The Part Where This Gets Useful
Here’s the thing nobody talks about when they talk about fishing algorithms: the logic isn’t actually about fishing.
Barometric pressure dropping fast → fish stop biting. They sense pressure changes before they happen. They adjust behavior based on leading indicators, not current conditions.
That’s… exactly how customer churn works.
By the time a customer is visibly unhappy (sending angry emails, not booking the QBR, going quiet in Slack), the pressure has already been dropping for weeks. The signals were there. You just weren’t looking at the right ones.
I run customer success for close to 3,000 accounts. There’s no way I’m reading every account manually, every week, without missing something. So I built the same thing I built for fishing: a composite scoring system that watches multiple signals and tells me when something needs attention.
The signals are different. Login frequency, support ticket volume, feature adoption rate, NPS trajectory, days since last outreach, renewal timeline. But the architecture is identical. Leading indicators in, composite score out, automated alert when the score drops below a threshold.
I didn’t learn this in a business school. I learned it because I was annoyed about bad fishing nights.
What the Score Actually Does
The customer health engine runs every morning. It writes a report. I open an HTML file (not a terminal scroll, not a markdown wall, an actual readable document) and I see which accounts are in red, which are yellow, which are green.
Red means: something’s wrong, look today. Yellow means: worth watching, add to the rotation. Green means: leave it alone, you’ll make it weird.
The script doesn’t tell me what’s wrong. It tells me where to look. The diagnosis is still human. The triage is automated.
This is the actual value of building systems: not that they replace judgment, but that they make sure judgment gets applied to the right things. Without the scoring engine, I’d spend my judgment on whichever account emailed me last. With it, I spend it on whichever account actually needs it.
The Absurdity Is the Point
People look at the fishing script and think it’s a funny eccentricity. “Look at the guy who automated his hobby.” And yeah, it is kind of funny.
But the fishing script is the customer health engine. Not metaphorically. Literally the same logic, same structure, same approach. The fishing score is a proof of concept that I built for personal reasons and then adapted for professional ones.
My best professional tools come from my dumbest personal obsessions. Every time.
The goal is never the tool. The goal is the thinking.
Tonight’s a 78
The fishing CLI just ran. Score is 78. Pressure’s been rising since noon, moon’s in quarter phase, wind is forecast at 8 mph and dropping after sunset.
I’m going fishing.
The customer health engine ran this morning. Two accounts in yellow. One I expected: they’ve been quiet since their main champion left. One I didn’t: adoption dropped 40% in two weeks and I hadn’t noticed.
I’ve got follow-ups scheduled for both on Monday.
Both systems are doing what they’re supposed to do: making sure I’m paying attention to the right things, so I can stop worrying about the rest.
Blake Bailey runs Bailey Business Ventures, an AI transformation consulting practice. He has built 1,300+ production automations, once spent six hours automating his fishing trips, and is deeply, personally offended by doing things manually.