How I retired early from learning coding at a young age

Written: 2017-07-12

My background of programming and how I eventually started and sold a company from it.

My story of an unorthodox way of building a tech business during the early .com boom where you can one-man-army it

I was brought into programming from an early age. My parents bought me a VIC20 when I was young. My first venture into computing was playing a game called Blitz. This time was before even ANSI. Basically, you hit the space bar to drop a bomb on the skyscrapers. I have no idea if I was any good or not, but my parents' encouragement convinced me that I was great at it.

Later, a Commodore 64. That was an era that was probably not so interesting to me because it was a cut down versions of arcade games (e.g. Yie Ar Kung-Fu).

I found real computing when the Commodore Amiga arrived. That thing was badass and had a history that I feel every geek should know. It also had an excellent soundboard that was used for many pop tracks at the time.

My first venture into coding was in one of my best friends house. He had a C64 or a VIC20, and I don't even recall. Either way, the language was BASIC. Clearly, in the 80s we had no internet, let alone BBS's or modems. What happened was that he had a computer magazine which had physically paper-printed code that you had to read and then type in (we were years away from OCR or having digital magazines). I had already become adept at typing from playing around with computers, so I punched it in. It didn't work, of course, because of all the typos (total absence of a linter). However, from what I recall, we eventually got it working. I barely remember what it did. However, we had programmed a computer. He encouraged me, and I did it.

Later, the Amiga era came. The most significant era of them all in my opinion (debatably SNES but that was before AAA video games were becoming a painful, inevitable, predictable termination to our indy fun).

Enter AMOS. Without looking it up (I'm flying), I recall it to be a wrapper around BASIC. Perhaps, in more modern words, it was a mini framework. I had this yellow elephant called "Trunky" at the time. All I wanted to do was make the yellow elephant move left and right with the keyboard. However, we didn't have Google back then. At the age of about 8 or so, I was a total novice. But I just about managed to move the elephant around with the keyboard after eventually stopping the CPU and disk drive going mental.

Btw, recently my mum uncovered an old sonnet I wrote in school at the age of 13 (while in year 9). Looking back, it seems I had already caught the bug. I interject with this as I believe it coincided with the above paragraph.

Rather quickly my interest in playing games turned into programming games.

I didn't really ever finish writing a game (either by myself or with my brother) to the extent of it being publishable. But we created a few. Distribution would have needed a far greater state of completion back then. We had no end goal in doing them, and it was just fun. I was also of a magpie nature, constantly trying out the forever evolving technologies.

I used all sorts of languages, from BASIC, C, C++, and later C based 3D engines. Storage was usually disk in some form of a database. I had a loose style, and a lot of my code was prototype code that just worked and where refactoring simply did not make sense as long as it was relatively legible.

I took a bit of a break from coding while I went to university to study computer science, ironically. I left after the first year with 12% attendance knowing that, not only was I wasting time learning what I already knew, a career behind a computer working for a bank wasn't for me. I think I had already envisaged my future "laziness" and entrepreneurial nature.

So, behind me I had near-2-decade's of knowledge of various languages (sometimes I would code 20 hours per day) - but more importantly - the knowledge that you can program a computer to do almost anything (well, anything "current", perhaps). Moreover, with this, I started writing (hands-on programming) the website for the new business in my bedroom a few days after my brother died. The house was full of extended relatives, and I often hid away upstairs as, being distraught, I chose to grieve without being ambushed with having to put on a brave face to strangers.

One thing that pushed me to do it, perhaps, was proof to my brother that we could have done it (together). It was only about 3 months prior that I told him about the idea of selling product online and (in fairness, drunkenly - mind you, we mostly only spoke after a drink) I professed how "easy" it would be for us. Sadly, that team wasn't to be.

Most other people I said it to doubted me except one of my friends who, in the pub, said: "I bet he'll bloody do it and all". That precise sentence still rings in my head when I think about it. If it wasn't for him and my brother I don't know if I would have made it.

Anyway, so there I was, coding upstairs on a single screen laptop in a brand new language. After about two weeks, I had a semi-working version of a website written in raw PHP. I needed to shortcut the process to maintain momentum, so I used some bits from some free shopping cart script back then to make it transactional and later had to go through the process of unpicking it out. It went live after about four weeks.

It didn't really work at first - I was trying to sell the wrong product and had a considerable business learning curve ahead.

About 6 months after, though, it worked. Really well.

This time the tech had to be serious. So I finally stopped jumping languages for the hell of it, and it stayed raw PHP. No framework, no packages - I didn't even use Composer. No 3rd party utilities. It cloc'd about 500,000 lines by the very end (I will come to that soon). I remained true to my ethos: If you use 3rd party tools, you are constraining yourself to the limits of what your competitors can do.
(Except maybe Zendesk, because writing a customer support system sucks...)

7 years later I sold the company and retired from it - at the age of 33, having been the single author of the entire codebase which commanded a barcoded 30,000 square foot warehouse of 120 pickers/packers, automated reordering, accounting, stock counting, HR, bulletproof security, an ugly but incredibly functional and idiot-proof CMS, handwritten conditional fraud check logic to handle thousands of orders per day without human intervention, and an obsessively conversion optimised front end. After selling out, I was replaced with a team of 12 programmers. The power of single-author code furthermore (business-owner wrote code) is often understated - where any change, any idea, can be deployed at my own level of risk at any time of the day. Even if that means waking up at 2am on a Monday going live with some crazy idea I had dreamt up. I did notice the other day that their JSON-LD was screwed on every page - they have thousands of product reviews and I noticed them not being leveraged in the SERPs. Perhaps their monitoring just needs a hot fix...

NB. That isn't a dig. I hope they are doing well and hope to see that company continue to strive and deliver the ROI the buyers deserve. It's just a bug.

I should consider that during this phase, I tried to launch another company selling product using the same technology. That didn't work. I underestimated the product-specific precision involved. I also overestimated my ability to lead and grow horizontally. I jumped the gun and failed there. Failing defines us.

I recall at one point when mobile was the new thing, and Google were always on my back to write a mobile site. I probably held off for about 6 months as I literally could not be bothered to write the CSS (and JS/markup of course). But one day I took a challenge from a Google employee where I would write the entire mobile version over a weekend. I worked probably 40 hours that weekend. Admittedly I think I drank a few bottles of wine to take the edge off, but by Sunday night it was live, and I dropped her an email to let her know. They were rather impressed, but so was I, as I grew the company's sales by about 5% in a weekend. Note: Clearly it wouldn't be as quick in these days now we no longer use markup in mobile 'apps'.

I did, however, (despite only having 6 staff in the office), need a great ops guy. That was initially my dad, who trusted my instincts, which to others were aggressively futurely sighted. I am thankful for that.

So, that's how coding paid off for me.

Or - maybe that is a over/mis-statement because the company was a hell of a lot more complicated than just coding my way to success, but even towards the end a good 20% of my time was spent doing it, and it eventually caught me up.

So how is any of this relevant to business? Because I built a tech business (pure-play e-commerce), which had a tech cost of $0. As opposed to millions (I witnessed that reality out after we bought out a lesser competitor later on). Because it had very little cost, it could be bootstrapped. So I didn't have to sell a single share. That's how coding is relevant to business in this rather rare example.

Early retirement sucks though. You may have to take my word for that. Because, I went to Ibiza for a year and went a bit loco, then... I ended up coding again.

Note: This blog speaks of a company that was sold in 2014 during the dot com boom; clearly the technological landscape has matured somewhat so do not consider my words to be of current advisory capacity.