A data room is not a checklist, and the difference costs you weeks

Every fund has a data room and every fund still loses weeks in diligence. Those two facts sit together uncomfortably enough that it is worth being precise about what a data room does, what it does not, and which of the two your process is actually leaning on.

Two different questions

A data room answers one question extremely well: what have we been sent, and who is allowed to look at it? Permissions, access logs, folder structure, watermarking — all of it is built around custody of files that already exist.

Diligence runs on a different question: what is still outstanding, who is it with, and how long has it been there?

Nothing in a folder tree can answer that. A folder knows that Accounts_FY24_FINAL_v3.pdf exists. It does not know that the FY23s are also required, that the FY22s were waived by the deal owner with a reason, that the bank mandate arrived inside a zip nobody has opened, or that item 19 came back unsigned and is therefore outstanding again despite a file sitting against it.

Why the gap gets papered over

The usual patch is a spreadsheet beside the data room. It is a reasonable instinct and it fails in a specific, predictable way.

You now have three records of the same truth: the checklist, the email thread where things were actually asked for, and the folder where they actually landed. Each is updated by a different person at a different time. The spreadsheet is accurate on the day it is written. By the end of the first week the thread has moved on, two files have arrived that nobody has filed, and one item has been answered in the body of an email rather than as an attachment — which means it is satisfied in reality and open on the list.

From that point the list is not a record of the deal. It is a record of someone's last attempt to keep up with the deal, and everyone on the team quietly knows it, which is why they ask each other rather than reading it.

The tell

There is one symptom that reliably shows the gap has opened, and it is worth watching for because counterparties notice it before you do.

Someone asks the borrower for something they already sent.

That is not an administrative slip. It is a credibility event, and the credibility being spent is yours, not the tool's. A finance director who has been asked twice for the same management accounts has learned something about how your firm runs a process, and it will come up the next time they are choosing between you and another lender.

The reason it happens is structural rather than careless: the person drafting the request cannot cheaply find out what has already arrived, because answering that would mean opening the folder, the thread and the spreadsheet and reconciling all three. So they ask. Asking is faster than checking, right up until it is not.

What a checklist has that an index does not

Four things, and each of them is the reason for a specific week of delay when it is absent.

A state per item, not per file. An item can be open, requested, supplied, proposed, in review, accepted, bounced back or waived. A file is just present or absent. Most of the interesting states in a diligence process — we have something against this but nobody has accepted it yet — have no representation in a folder at all.

An owner per item. Half the items on a real checklist are owed by someone inside your own building: Legal reviewing the facility agreement, Credit signing off the model. A data room has no concept of this, so internal work does not appear on the list at all, and the delay it causes is invisible.

A many-to-many link between evidence and requirement. One management accounts PDF can satisfy three requirements at different page ranges. One requirement can need three documents. Folders model neither, which is why "which items does this file actually close?" is a question a human answers by reading.

A record of who accepted what, on the basis of what. "Item 14 was accepted by Anna on 3 September, on the basis of pages 3 to 5 of this file" is the sentence an investment committee needs and a folder cannot produce. It matters most in the one case nobody plans for: when a deal goes wrong and the question becomes who decided this was fine?

Waivers are the clearest example

Every real deal waives something. The board approval that turned out to be below the delegation limit, the third year of accounts for a company that is two years old.

In a folder-shaped process a waived item looks exactly like a satisfied one: nothing is missing, so nothing shows up as missing. That is the failure mode worth being frightened of, because it is silent. A waived item should read as waived, with a name and a reason against it, and it should read that way in the pack that goes to committee — not as quietly complete.

What to do about it

Keep the data room. It is good at custody, and replacing it is not the point.

Put the checklist somewhere that can hold state, ownership and evidence links, and make the data room one of the places evidence arrives from rather than the system of record for whether the deal is done. Then the question everyone actually asks — where is this deal and who is it waiting on? — has an answer that does not require anyone to open three tabs and reconcile them by hand.

The measure worth watching is days from opening a deal to IC-ready. It is the only one that captures both halves of the delay: what the counterparty owes you, and what your own reviewers are sitting on.