Run the query
Take any CRM that has been in place for three years and count the custom fields. Then count how many have been written to in the last twelve months.
The second number is usually somewhere between a half and two thirds of the first. The remainder are not neutral. Each one is on a form somebody has to look past, in an import somebody has to map, and in an export nobody reads.
Every one was a good idea on the day
That is what makes them interesting rather than embarrassing. Nobody adds a field frivolously. Each was added because a real person had a real question — a campaign that needed tracking, a quarter where the board wanted segment reported a particular way, a process that has since been replaced.
So the dead fields are a fairly precise record of what the company cared about, and when it stopped. Sorted by date of last write, they read like a diary of abandoned initiatives.
The cost is attention, not storage
The instinct is to say they are harmless. Storage is free and an unused field takes no space.
But every field on a form is a small tax on the person filling it in, and the tax is paid at the worst moment — while somebody is trying to record something real about a customer. Twenty optional fields is why records get filled in badly, and a record filled in badly is why the reporting is distrusted, which is why somebody eventually builds a spreadsheet.
What to do about it
Not much, and deliberately. Delete the ones that are obviously dead, after checking nothing reports on them. Archive rather than destroy, because somebody will ask.
The more useful move is to make adding one slightly harder than it currently is. Not a committee — just a named owner and a question about what decision the field will inform. Most requests do not survive that question, which is the point.
And run the query annually. It takes a minute and it is the cheapest audit of organisational attention you will ever do.