File formats

What to Do With Files Nobody Can Explain

Every individual step was sensible. None came with a note saying what it was, because writing that note was nobody job at the time.

Robert Smith 10 September 2026 6 min read

Inheriting files nobody can explain is an inventory problem before it is a technical one. A drive appears. A mailbox export nobody remembers making. A folder from somebody who left three years ago. Nobody knows what is in it or whether it matters. The person who could have said is not available.

This happens constantly and almost nothing is written about it, because there is no single product that answers it. What follows is a method rather than a solution.

Establish What the Files Are Before You Decide

The instinct is to start deciding. Keep this, bin that, convert the rest. Every one of those decisions depends on facts nobody has yet.

Questions People Start With

Can we delete this
Should we convert it
Where should it live
Who owns it now

Questions That Come First

What is actually in here
What can be read at all
What period does it cover
Is any of it duplicated elsewhere

The right column is an afternoon of work with no judgement in it. The left column is a decision somebody senior has to make. It becomes easy once the right column is answered.

Separating those two is most of the job. The establishing work needs somebody methodical and nothing else. The deciding work needs authority. Collapsing them means whoever opened the drive is quietly making retention decisions they were never asked to make.

Building an Inventory of Inherited Files

The aim is a page of facts, not a catalogue of every file.

Group by what the files really are. Not by extension. Read the opening bytes where the names look unreliable, which in an inherited folder they usually are.
Record the date range. Oldest and newest, taken from inside the files where possible. Folder dates often reflect when somebody copied everything rather than when anything was made.
Note the size distribution. One four gigabyte file and nine thousand small ones are different problems. The shape tells you which you have.
Open one of each kind. Not all of them. One. Enough to say whether that group is readable and roughly what it holds.
List what would not open. By format and count. This is a finding rather than an admission.

Do all of this on a copy. Take one, work on it, leave the original untouched and unopened. Some formats change on being opened. Where the collection might ever matter formally, that is the difference between evidence and a file somebody has been in.

What You Are Likely to Find

Inherited collections have a recognisable shape. Knowing it in advance saves working it out twice.

What turns up	What it usually means
Several copies of the same thing	Somebody backed up by copying, repeatedly
A large mail store from a leaver	An export made to get under a quota
Files in a format nothing opens	Software the organisation stopped using
A folder called final, plus one called final2	Ordinary work. The newest is not always right
An encrypted archive	A problem, since the password left with the person
Personal material mixed in	Somebody's actual life, in a work folder

The last two rows need naming early. An encrypted file nobody can open is worth flagging on day one rather than at the end, because somebody may still know. Personal material needs a decision from somebody who can make it. Quietly reading through it is not that decision.

Judging Which Inherited Files Are Worth Keeping

Once the inventory exists, the decision is somebody else's. What you can do is frame it usefully.

Business record Keep, with a note on whyContracts, decisions, correspondence a client could ask about. This is the part with an owner and usually a required period
Working material Usually not worth keepingDrafts, intermediate files, copies of things held elsewhere. Bulky. Rarely asked for once the work finished. Held elsewhere as often as not
Unidentifiable Keep until somebody looksCheap to store and impossible to recreate. The one category where doing nothing is defensible

Three groups rather than a file by file judgement. Anything finer than this takes longer than the material is worth and produces a decision nobody can apply consistently.

Ask about holds before deleting anything. An inherited collection is exactly the kind of material that turns out to be relevant to something. A hold overrides ordinary retention. One question to the right person costs nothing and settles it.

Converting Inherited Files, If At All

The temptation is to modernise everything at once. It is usually the wrong move and always the wrong first move.

Converting too early replaces one unclear collection with another. It also destroys the record of where you began. Something is always dropped. Dropping it before anybody has decided what matters is a decision made by accident.

Where converting is right, keep the source. The original stays untouched alongside whatever comes out. Then the conversion can be redone when somebody notices it lost something, which happens more often than not.

What You Actually Produce at the End

The output is a document rather than a tidied folder.

What is here
  roughly how much, in what formats, covering what years

What was readable
  which groups opened, what they appear to contain

What was not
  which formats, how many files, what was tried

What I did
  copied, inventoried, nothing deleted, nothing converted

What needs a decision
  the encrypted archive, the personal material,
  whether any of it is under a hold

The notes outlast the decision. Whatever gets decided about the contents, the next person to meet this material inherits the folder plus your document. Without it they start exactly where you did, which is what happened to you.

Why This Keeps Happening

Nobody sets out to leave an unexplained folder behind. It accumulates because every individual step was sensible.

Somebody exported a mailbox to free space. Somebody copied a drive before replacing a laptop. Somebody kept a leaver's files in case they were needed. Each act was reasonable and none came with a note saying what it was or how long it should stay, because writing that note was nobody's job at the time.

The only real prevention is a note beside the file. A plain text file saying what this is, where it came from and who to ask. It takes two minutes at the moment of creation and it is the one thing that would have saved you the afternoon you just spent.

This is a working method rather than a standard. Where a legal hold, a regulator or a contract applies, that instruction governs and nothing here does.

Questions People Ask

6 questions, answered in full below.

Where do I start with a folder nobody can explain?

With an inventory rather than a decision. Count what is there, group it by what the files actually are and note the date range. Everything anybody will ask you depends on facts you do not have yet. Gathering them takes an afternoon.

Should I convert everything to a modern format first?

No. Converting first turns one unknown collection into a different unknown collection. It also wipes out the evidence of where you began. Establish first, convert deliberately afterwards if at all.

What if I cannot open some of the files?

Record which ones and move on. Not opening is itself a result worth writing down. It is useful to whoever decides next. Stopping the whole exercise to solve one format is how these projects stall permanently.

Is it safe to delete obvious rubbish?

Not yet. Possibly not at all. Ask whether anything might be under a hold before deleting anything from an inherited collection. Obvious rubbish is also cheap to keep, so the case for removing it early is weaker than it feels.

How long should this take?

The inventory is an afternoon for most collections. What takes longer is the decision about what to keep. That belongs to somebody else

What do I hand over at the end?

A short note covering what is there, what opened, what did not and what you did. The files were already present. The notes are what you added. They stop the next person starting from nothing.

Sources

Where the figures and behaviour described above were checked.