Day 990 - The Financial Anatomy of a Knowledge-Work Business Collapse
Lab notes, Day 990
A few weeks ago, I told you my training business died in six months. That was the headline. This is the autopsy.
I want to be specific. Not therapeutic-specific — I’m not interested in processing feelings here. Financially specific. Because the numbers tell a story that is more useful to you than the narrative I was telling myself while it was happening.
How it started
March 13, 2020. New Zealand was four days from its first COVID lockdown. I was on a contract at a major Auckland tech firm, and I was frustrated. The same mistakes, everywhere. Bad naming. Tangled dependencies. Tests that tested nothing. Developers who were skilled but had never been shown a better way.
So I opened a Slack channel. I called it #olafs-daily-tips. Each day, one post: a clean code technique, a SOLID principle, a pattern for handling legacy code without breaking it. I thought I might get ten or twenty followers if I was lucky.
At its peak, the channel had just shy of 350.
When that contract ended in May 2020, I had a decision. I could file the tips away and go back to being a senior developer. Or I could find out if there was a business in what I’d been doing for free.
I set up a website. I kept writing. In the end: over 300 posts on software craft, covering everything from unit testing to clean architecture to component design principles. The kind of material I’d been wanting to exist for twenty years.
Word spread. By March 2021 — almost exactly a year after the Slack channel opened — I had my first paying client: a major NZ fintech. One workshop on TDD. A few thousand dollars. Proof of concept.
The good years
FY22 was the peak. Total revenue: $222,096.
The headline number flatters the underlying business a little. About $200,000 of it came from a single six-month embedded training contract with a large NZ tech company — essentially a coaching residency, working alongside their development teams every day. Strip that out and the standalone workshop business was around $22,000 that year.
But still. The business was real. The pipeline was building. I had:
A major NZ fintech (repeat client, growing engagement)
A Big Four bank
A global payments network — a genuine showcase client, the kind that makes other prospects take you seriously
An Australian software firm, trained in Bath, UK, via video call, up to midnight my time
That last one felt significant. Export dollars. International reach. I was no longer just a local training operation.
FY23 brought in $109,000. Down 51% from peak — but I didn’t read it that way at the time. The large embedded contract hadn’t repeated, which explained most of the gap. The global payments network was a new client. The Australian firm was back. The major NZ fintech was still booking.
Every business has down years. I’d had a down year. That was my read.
I pushed harder. LinkedIn posts, daily tips in an already crowded market, hoping to be spotted by a tech leader with budget. I built an online course on TDD using Teachable. I kept doing the midnight calls for international clients. I sent a monthly newsletter to a list of 43 people — every client manager I’d worked with, a handful of prospects, a few aspiring engineering managers. People with budget authority. People who had seen the work firsthand. I offered 30%, 33%, sometimes 50% off workshops and seminar bundles.
Six students enrolled in the Teachable course. Total.
Still, I thought: keep building, keep the quality high, the market will come back.
The six months
It didn’t come back.
October 2023 to March 2024. That’s the window. That’s when it became undeniable.
The newsletter kept going out. The discount offers kept going out. The 43 people on that list — the right people, the exact right people — read them and did nothing. Not a negotiation, not a “not right now”, not even a polite decline. Silence.
FY24 revenue: $47,000. Down 57% from FY23. Down 79% from peak.
By March 2024, I knew. Not that the business was struggling — I’d known that for a while. That the game was up.
