Systems / Work / Design / Operations
I like systems, but I do not think everything needs one. My instinct around messy information is usually to organise it. Give things names. Decide where they live. Make the relationship between one thing and the next easier to see. That instinct is useful until organisation becomes an end in itself.
A system earns its place when it removes a question that people should not have to solve from the beginning every time. Where does this information live? Which version is current? Who owns the decision? What happens after this step? What do we measure? If the same ambiguity keeps returning, there is probably something worth structuring.
But structure can create its own kind of ambiguity. A new dashboard might summarise information beautifully while creating another place that needs to be updated. A template can make output more consistent while forcing every problem into the same shape. Documentation can become extremely complete and still fail the person who opens it because the next action remains unclear.
That is the difference I keep returning to between recording information and changing what somebody has to decide next. Documentation can tell me what exists.
A useful system should also make it easier to know what happens because it exists. This is why I am more suspicious now of systems that look impressive from the outside but require constant explanation from the person who built them. If nobody can use the structure without remembering the structure, we may have simply moved the problem somewhere more organised.
The simplest test I know is still the most useful one:
Does this make the next decision easier?
If it does, keep it. If it creates three new decisions before anyone can make the original one, it probably needs another pass.