There was a bug in the billing system that had existed for three years.
It wasn't a serious bug. Under certain conditions — a specific combination of timezone, currency, and subscription type — invoices would show the wrong date. Not the wrong amount, just the wrong date. Off by one day.
Customers occasionally noticed. Support would apologise, manually correct it, and move on. The bug was documented. Everyone knew about it.
When I finally looked at the code, the fix took twenty minutes. A single function was using the wrong timezone conversion. The test was missing because the original author hadn't considered that edge case. Understandable. Easy to miss.
What wasn't easy to understand was why it had taken three years.
The bug had been reported internally at least a dozen times. It had been added to the backlog, prioritised, deprioritised, moved between teams, and eventually forgotten. Every few months, someone new would discover it, file a ticket, and the cycle would repeat.
The fix was trivial. The organisational friction was not.
The billing system was owned by a team that had been reorganised twice. The original authors had left. The current owners were focused on a migration project and weren't taking on "minor" bugs. The product team didn't consider it a priority because it didn't affect revenue. Support had learned to work around it.
Everyone had a reasonable explanation for why it wasn't their job to fix it.
This is how small problems become permanent. Not because they're hard, but because fixing them requires crossing boundaries that nobody wants to cross.
The bug wasn't a technical failure. It was an organisational one. The system had evolved to make certain kinds of work invisible. Maintenance, cleanup, small fixes — these didn't show up on roadmaps or OKRs. They didn't get celebrated. They just happened, or they didn't.
I fixed the bug because I happened to be looking at that code for another reason. It was accidental. If I hadn't been there, it might have lasted another three years.
Afterwards, I started noticing similar patterns everywhere. Small problems that everyone knew about but nobody fixed. Not because people were lazy, but because the organisation had made fixing them irrational. The incentives pointed elsewhere.
The lesson wasn't about bugs. It was about what organisations optimise for, and what they quietly neglect.
If you want to understand how a company actually works, don't look at the roadmap. Look at the bugs that never get fixed.