Chapter Three · Story · 3 min
The Night It Actually Clicked
For a while I thought the discovery was speed. That is the obvious thing to notice when you can make something useful in an evening that once needed a project plan, a budget line and three meetings before anybody was allowed to touch a keyboard.
For a while I thought the discovery was speed. That is the obvious thing to notice when you can make something useful in an evening that once needed a project plan, a budget line and three meetings before anybody was allowed to touch a keyboard. Speed is intoxicating. It is also a very good way to miss the point.
The moment that changed the way I saw it came at an adventure centre. I sat down with the owner expecting to talk about a booking system. He showed me the first system. “This is the bit we use for bookings,” he said. “It isn't great, but it does most of what we want.” It did not take many questions before “most” began doing quite a lot of work in that sentence.
The booking lived there. Then, because that system did not handle what happened next, the details were entered again into another system that had been developed with somebody overseas at considerable cost. That created the programme for the group staying at the centre. Then another tool was used to schedule the team. Information was pulled from different places to communicate with the customer. Eventually more of it was entered into the finance package so an invoice could be produced.
Booking. Something in the middle. Invoice.
Except the “something in the middle” was where the business actually lived, and it had been split across systems that did not talk to one another. The staff had become the integration layer. Human APIs, except with coffee and occasionally bad language.
At one point I asked for a login so I could review one of the systems properly while we explored a better way of working. The owner explained that the user-management bit did not work properly. Every time somebody joined or left, he had to contact support and ask them to sort it.
That detail stayed with me because it was so ordinary. Nobody had called a crisis meeting about it. Nobody had a transformation programme to fix it. They had simply worked around it for long enough that the workaround had become part of the job. And suddenly I could see friction everywhere. Not dramatic failure. Friction. Copy this. Re-key that. Phone support. Check the spreadsheet. Explain the exception. Apologise to the customer. Remember the thing the system does not remember.
The mechanic came soon afterwards and made the pattern impossible to unsee. Different industry, different language, different work. Yet underneath it sat the same shape, a booking, something important in the middle, and an invoice at the end. The middle changed. The human need did not.
That is when it stopped being, in my head, a faster way to build software. It became a way of seeing the work before deciding what, if anything, should be built.
What struck me afterwards was how little of the conversation had been about replacing a booking system. If I had walked in asking which features the owner wanted in a new platform, I would have left with a respectable list and almost certainly the wrong understanding. The useful part came from following the work. Where does this information go next? Who touches it? What do they have to remember? What happens when somebody changes their mind? Each answer opened another door, and behind nearly every door was a person doing a small piece of unpaid integration work on behalf of the systems around them.
That matters because the lost time was not sitting neatly in one place waiting to be measured. It was scattered through the day in two-minute jobs. Copy a name. Check a date. Ring support. Re-enter a number. Send an email because the systems could not speak to one another. None of those jobs was large enough to justify a project on its own, which is exactly why they had survived. Put them together, though, and you could see the business spending a surprising amount of human attention simply keeping its technology joined up.
I am wary of making that sound more profound than it felt at the time. There was no whiteboard moment where Neil and I circled the word 'pattern' and congratulated ourselves. It was more irritating than that. I simply started noticing the same thing everywhere. Once you notice a loose floorboard you somehow step on it every time you cross the room. The world had not suddenly filled with friction. I had finally learned to hear the squeak.