Which file is current?

A cost-center list gets emailed around, edited in three places at once, and saved back with a new suffix each time. Eventually someone has to guess which copy is right. Here is why that guess is more common than most organisations admit.

Open the shared drive folder for almost any reference list, cost centers, product mappings, region codes, and you will find more than one file. mapping_2024.xlsx, mapping_2024_final.xlsx, mapping_2024_final_v2.xlsx, mapping_JP_edit.xlsx. Someone, at some point, has to decide which one is real. Usually that someone is whoever opens the folder next, and they decide by guessing: newest timestamp, longest filename, or just asking around.

How it gets there

No one sets out to create five versions of a file. It happens gradually. A controller updates a cost-center list and emails it to a colleague for a second opinion. The colleague makes a small correction and saves a new copy, because overwriting someone else's file feels risky. A third person, working from an older email, edits their own copy without knowing the other two exist. None of these steps is a mistake by itself. The mistake is a format, the spreadsheet, that has no idea any of this is happening. It cannot tell you that two other copies exist, let alone which one has the change that matters.

Why "just use SharePoint" doesn't fix it

The common answer is to put the file in one shared location instead of emailing it around. That helps with discovery, there is now one place to look, but it does not answer the actual question. A shared file can still be edited by five people across a day, saved over itself repeatedly, and reverted by an autosave nobody asked for. Version history in these tools exists, but it is a list of timestamps and file diffs, not a record of who changed which value, why, or whether anyone checked it. Finding out what changed still means opening old versions and comparing cells by eye.

What the question is really asking

"Which file is current" is rarely just about the file. It is really three separate questions bundled into one: which values are correct right now, who is responsible for them being correct, and can I trust a number enough to put it in a report. A single authoritative copy answers the first question. It does nothing for the second and third. Those need a role attached to the data (who is allowed to change this table) and a record attached to every change (who changed this value, and did someone check it). A spreadsheet, no matter where it lives, has neither.

What a workable answer looks like

The organisations that stop losing time to this problem tend to do three unglamorous things. They put each list in exactly one place, so there is never a question of which file to open. They attach a role to it, so it is clear who may change it and who may only read it. And they keep every change as a new version rather than an overwrite, so "what changed and when" is answered by looking, not by asking around. None of this requires sophisticated tooling. It requires treating a list of a few hundred rows with the same basic discipline as a codebase: one source, tracked changes, a review step before something counts as final.

The alternative, in most companies, is not a system at all. It is a folder, a naming convention that nobody enforces, and a habit of asking "is this the latest one?" in a Teams message before every report. That question, asked often enough, is itself the cost.

---

Bifrost is our attempt at making these basics practical. It is currently in pilot — you can read more or try the demo from the homepage.

All articles