The way I define technical debt is something that risks wasting someone's time. If you decide to make a change quickly to get something working and create a mess in the process, then it's possible that that mess will waste someone's time when it comes to making other changes. Note this is separate from a mess that can cause bugs. For example a significant lack of unit tests I would classify as a business or data risk - anything that risks losing customers should be given higher priority than technical debt.