The Data Room: How Diligence Information Is Controlled

A data room is the controlled place where a seller discloses information to a buyer during diligence. Access is granted person by person and folder by folder, disclosure is staged, sensitive material is redacted or restricted, and every view is logged. Control is the purpose. Storage is incidental, and a room used only as storage has been misunderstood by whoever built it.

Almost everybody who hears the phrase for the first time pictures a filing system. A place where the documents go. Somewhere on a server, with folders, and a login. The filing-system picture is not wrong so much as it is upside down. The documents are the easy part. The hard part, and the reason the whole apparatus exists, is that a seller has to place its most sensitive information in the hands of a buyer. The buyer may, at any point in the next several weeks, decide not to buy anything at all and go back to being what it was before. On this transaction, what it was before is a business selling to some of the same customers.

So the useful way to hold the idea is this. A data room is not a place where information is kept. A data room is a set of decisions about who may see what, in what order, in what form, with what recorded afterwards. The decisions are the substance. The software that carries them out is the part that has a login screen, and it is the least interesting layer of the whole thing.

Harivansh Packaging Limited, an invented listed maker of rigid and flexible packaging, reports revenue of Rs 3,180 crore. Harivansh Packaging is buying Sundarban Polymers Private Limited, an invented unlisted maker of flexible packaging films with revenue of Rs 880 crore and earnings before interest, tax, depreciation and amortisation (EBITDA) of Rs 132 crore, a margin of 15.0 per cent. Devyani Kulkarni is the chief financial officer of Harivansh Packaging Limited and Ashwin Rege leads its transaction team. The transaction's sequence runs approach and confidentiality, indicative offer and term sheet, confirmatory diligenceThe detailed examination a buyer runs over a business once a shape has been agreed, to confirm what it has been told before it commits to buy., documentation, signing, the conditions period, completion. The room is built for the third milestone in that sequence.

Two things are assumed here. Why the purchase is being attempted at all was settled before this sequence began. And the shape of the deal was agreed at the term sheetA short document setting out the agreed shape of a transaction, before the long agreements are drafted. A term sheet fixes what the parties have already agreed rather than everything they will eventually sign., the milestone immediately before confirmatory diligence. A room is normally opened after that shape exists. The consequence is larger than the fact sounds: the buyer walking into the room already has an offer on the table, and the room is where that offer gets confirmed, adjusted or withdrawn.

What is a data room actually for?

Two things have to be true at the same time, and they pull hard against each other. A buyer cannot commit hundreds of crores to a business it has not examined, so it must be shown enough to be sure of what it is buying. A seller cannot hand over everything it knows to a buyer who might not buy. On the day that buyer walks away, the buyer keeps everything it read. Both halves of that sentence describe a real constraint, and the room exists to hold them apart rather than to choose between them.

Take it out of finance for a moment. The shape is completely ordinary. A shopkeeper is selling a shop. A buyer turns up and asks to see the takings. The shopkeeper does not refuse. A buyer who cannot see the takings will not pay for them. But the shopkeeper does not leave the takings book on the counter either. The takings book stays in the back room, it is shown to the buyer while the shopkeeper is standing there, and it is put away afterwards. If the buyer is a stranger from another town, that is one conversation. If the buyer runs the shop two streets down, it is a very different one, and the shopkeeper knows it before anyone explains it.

The finance version has the same two halves and more people. Harivansh Packaging Limited has to be sure the Rs 132 crore of EBITDA it is paying for is real, repeating and unencumbered. Somebody on its side therefore reads the trading record, the contracts, the borrowings and the working capital position. Sundarban Polymers Private Limited has to allow that reading. But Sundarban Polymers also has to live with the possibility that the transaction stops. Nothing forces it to complete. Conditions can fail, funding can move, the parties can simply not agree. On the morning after a transaction stops, everything the buyer read is still inside the buyer.

Which is why control, and not storage, is the purpose. A room built as a control has four working parts: who may open which folder, when each folder opens, what has been taken out of a document before it opens, and what record is kept of every opening. Take those four away and what remains is a shared drive. A room with all four is a way of disclosing something without giving it away, and that is a genuinely difficult thing to build.

Who gets access, and to what?

Here is the first thing that surprises people who have only heard the phrase. Access is not granted to the buyer. Access is granted person by person and folder by folder. Every cell in the grid of people against folders is a separate decision somebody made on purpose. Harivansh Packaging Limited as an entity does not have a login. Named individuals do, and each of those individuals sees a different room.

Why go to that trouble? Because a transaction of this size carries a surprising number of people, and they are looking for entirely different things. Ashwin Rege and the transaction team are testing the shape already agreed. The buyer's financial advisers are testing the trading record behind the Rs 132 crore of EBITDA. The buyer's tax advisers are reading tax computations and assessments. The buyer's legal advisers are reading contracts and the consents they may require. The lender underwriting the new borrowing is putting money in against the same business, so it does its own reading. Each of those has a reason to look at something. None of them has a reason to look at everything.

Put a name to the failure that follows from ignoring this. If the buyer's tax adviser is handed the whole room, that adviser is now a person holding the target's customer contracts. Nobody did anything wrong. Nobody misused anything. But the number of human beings holding commercially sensitive information about Sundarban Polymers Private Limited just went up by one, for no reason, and every one of those people is a place the information can go. An unnecessary viewer is not a neutral administrative fact: it is an additional copy of the information sitting in an additional head.

The household version is immediate. A large wedding is being organised and a dozen people are being let into the house. The caterer needs the kitchen and the guest count. The photographer needs the hall and the schedule. The decorator needs the terrace. Not one of them needs the cupboard where the jewellery is kept, and nobody in the house thinks of that as an insult to the caterer. Leaving the jewellery cupboard out is simply the ordinary practice of giving each person the access their job requires and no more.

So the access list is a decision, not an administrative step. On this transaction it is drawn up by Ashwin Rege's team against the folder structure and approved by Devyani Kulkarni before anybody logs in, and it is revisited each time somebody new joins the work. Revisiting the list matters more than it sounds. Transaction teams change during a long process, people are added to solve a specific problem, and the request that arrives is almost always to give somebody access, rather than to give somebody access to these four folders. The discipline is answering the second question when only the first one was asked.

Why is disclosure staged rather than opened all at once?

Because of one property that makes disclosure different from almost every other decision on a transaction. Disclosure cannot be reversed, so a seller gives up at each stage only what that stage requires, and staging is the mechanism that stops a walk-away from becoming a giveaway. Every other term on a deal can be renegotiated. A price can move, a condition can be dropped, a date can shift. A document that has been read has been read.

Think about what that does to the seller's incentives. If Sundarban Polymers Private Limited put everything into the room on day one, it would have handed over the full commercial picture of the business in exchange for an offer that is, at that moment, still indicative. The buyer has not yet committed to anything it cannot walk away from. The seller has committed everything. No seller would choose that negotiating position, and the reason is not distrust. The question is what happens if the transaction stops for reasons neither side controls.

So the room opens in stages, and each stage is tied to the decision the buyer has to make next. An early stage carries what a buyer needs in order to firm up the offer it has already made: the trading record behind the Rs 132 crore of EBITDA, the borrowings that sit behind Sundarban Polymers Private Limited's net debt of Rs 180 crore, the working capital position that will later be measured against the working capital pegA level of working capital written into the agreement as the normal one, against which the actual level at completion is measured so the price can be adjusted. of Rs 96 crore. A later stage carries what a buyer needs in order to sign: the contracts with the two counterpartiesThe businesses on the other side of a contract. Here, the two whose agreements change hands and whose consent is needed. whose consents become conditions, and everything else the paper will have to describe.

Notice the logic of that second stage. A buyer cannot sensibly accept a condition it has not read the clause behind. If completion depends on two counterparties consenting, the buyer needs to see what those contracts actually say before it agrees that their consent is a condition rather than a problem. The counterparty contract is a document a seller would rather not release early, and a document the buyer cannot sign without. Staging is how both of those are true at once.

Staging has a cost and it is worth naming honestly. Staging is slower. Buyers dislike it and say so. A team that has to wait for a folder is a team burning days. Sellers who overstage lose credibility and momentum, and a room that opens too little too late reads as evasion whether or not it is. The judgement being made is not how little can be shown, but what this next decision actually requires. Because it is a judgement, it sits with the people running the transaction rather than with whoever administers the software.

What is redacted, and why does a buyer who is also a competitor change the answer?

Now the part that makes this particular room harder to build than most. The record on this transaction carries a plain fact that decides the design: Sundarban Polymers Private Limited sells to some of the same customers as Harivansh Packaging Limited. The overlap is why the transaction exists at all. Overlapping customers are what make the two businesses worth combining. The overlap is also why the room cannot be built as a simple set of open folders.

Work through what customer level pricing actually is in that situation. Customer level pricing is a list of what the target charges each named customer. To a buyer that has decided to complete, it is diligence: it shows whether the Rs 132 crore of EBITDA rests on prices that hold. To a buyer that walks away, the very same file is a rival's price list, obtained at no cost, with the customers named, and no undertaking of confidentiality unwinds the fact that it has been read. Both readings sit on one document, and which one turns out to be true is not known while the room is open.

There are two ordinary answers, and rooms of this kind usually use both. The first is redactionBlacking out parts of a document before it is shown, so the rest can be read while the sensitive part does not cross.: the document goes into the room with the customer names removed, or the per unit prices removed, so the shape of the trading can be assessed without the identities and the prices travelling together. A buyer can still see that the largest customer is a fifth of revenue without learning which business that customer is and what it pays.

The second is a restricted groupA small set of named people, often from the buyer's advisers rather than its own staff, permitted to see material nobody else on the transaction may see. The group reports conclusions and does not pass the underlying file on., sometimes called a clean team. A small number of named people, usually people who are not going to be running the business afterwards, see the unredacted file. The restricted group then reports conclusions to the transaction team, and only conclusions. The team learns that the prices in the largest three contracts are stable and have not been discounted to hold volume. The team does not learn what those prices are. Hearing a conclusion is a genuinely different thing from reading the file, and the difference is the entire point of the arrangement.

One boundary must be stated flatly. Whether either of those arrangements is permitted between two businesses that share customers is a question for the competition authority. What may lawfully pass between competitors before a transaction completes is set by the authority that supervises competition, and the design of a room does not override it. The commercial mechanism and the reason it exists are one question. Whether the arrangement is permitted is another, and the permission is somebody else's to give.

What does the access log record, and who reads it later?

Every room of this kind keeps an access logThe automatic record a room keeps of which named person opened which document, and when.. The log records which document was opened, by which named person and at what time, without anybody having to remember to do it. The log is produced by the software, it accumulates quietly through the whole diligence period, and almost nobody looks at it while the transaction is live.

Then, some months after completion, somebody looks at it very carefully indeed. The log is the evidence of what was disclosed and to whom, so it is a governance record rather than a feature of the software. The evidence becomes the question the moment a buyer says it was never told something. That is the whole reason it belongs in an account of how a transaction is run.

The log answers two different questions for two different sides. For the seller, it answers: was this disclosed? The document sat in folder four, it was opened on three occasions by two named people on the buyer's side, and here are the timestamps. For the buyer, it answers a question about itself: who inside the organisation holds this? When a transaction stops, or when a restricted arrangement has to be shown to have worked, the buyer needs to be able to say precisely which individuals read which files. Without a log that is a matter of memory, and memory eight months later is not evidence of anything.

The log gets overread, so its limits need stating precisely. The log records that a document was opened. The log does not record that the document was read, that it was understood, or that its significance was appreciated by the person who opened it. A file opened for eleven seconds and a file studied for two hours look the same in most logs. So the log settles what was made available and to whom, and it does not settle what anybody knew. The gap between those two is precisely the territory the agreement will later have to deal with, and how it deals with it belongs to the sequence where the paper is read.

How do questions and answers run beside the documents?

A room is documents. A diligence process is questions. Documents and questions are not the same activity and neither one substitutes for the other. Every properly run room therefore has a second machine bolted to the side of it: the question logThe single numbered list of questions the buyer has asked and the answers the seller has given, kept in writing so both sides can point at the same text later..

The question log works in the plainest way imaginable. The buyer's people submit questions, numbered, into one channel. The seller's people answer them, in writing, against the same numbers. Nobody answers a diligence question by telephone, in a meeting, or on the way out of one. Questions go through a single channel and are answered in writing. The alternative is an answer given in a corridor that turns into a disputed disclosure eight months later.

The reason is not distrust between the parties. The reason is that human recollection of a conversation is genuinely unreliable, and that two honest people can walk out of the same meeting holding different versions of what was said. In writing, the answer exists as text. Both sides have the same text. If the answer was wrong, it is visibly wrong and it is attributable. If it was right, it protects the person who gave it just as much as the person who received it. Written answers help the seller at least as much as they help the buyer. Sellers insist on the channel rather than merely tolerating it.

There is a second use for the question log that is easy to miss. The question log tells the seller where the buyer's worry sits. Thirty questions about one customer relationship is a signal, and Devyani Kulkarni's side of the transaction learns something from it: this is where the price pressure is going to come from. A seller reading only the documents it uploaded knows what it disclosed. A seller reading the question log knows where the buyer is actually digging, and that is the more useful knowledge while a negotiation is still running.

And the pairing matters in the other direction too. Buyers who work only from documents end up assessing what happens to be in the room. Buyers who work from a question list assess the business, and use the room to answer it. The second is diligence. The first is reading.

What does a data room not do?

Three things, and each of them is regularly expected of it by somebody.

A data room does not analyse. A room holds a trading record; it does not say whether the Rs 132 crore of EBITDA repeats. A room holds contracts; it does not say which clause matters. Every conclusion drawn during diligence is drawn by a person, and the room's contribution is to have put the right document in front of the right person. Putting the right document in front of the right person is a real contribution, and it is not the same as insight.

A room does not tell a buyer what to look for. The room reflects what the seller decided to put in it, arranged the way the seller decided to arrange it. A buyer who lets the folder structure set its agenda has allowed the seller to write the diligence plan. The question channel exists for that practical reason: it is the buyer's agenda pushing back against the seller's arrangement.

And the third one is the trap, so it gets said flatly. Volume is not disclosure, and a room with forty thousand documents and no indexThe map of a room: a structured list of every folder and document, so a reader can tell what exists without opening anything. is a way of disclosing something while making it very unlikely that anybody finds it. The two are independent properties. A large room can disclose very little and a small one can disclose almost everything. The reason is that disclosurePutting information in front of the other side in a way it can actually take in. Making a document available is the first half of it; making it findable is the second. is a property of findability and not of quantity.

Ordinary life makes the same point. If somebody asks for the electricity bill and is handed a sealed carton containing every paper the household has kept for nine years, the bill is in there. The bill has, in a narrow sense, been provided. Nobody would say it was shown. The carton is a way of complying with the request while defeating its purpose, and a room without an index is the same carton with a login screen.

The honest position, then, is that a data room is a necessary piece of infrastructure and never a sufficient one. A data room makes controlled disclosure possible. A room does not make disclosure good, it does not make diligence competent, and its size tells a reader nothing at all about either.

How is this room built, folder by folder?

Put the pieces together on the transaction itself. Harivansh Packaging Limited, with revenue of Rs 3,180 crore, is buying Sundarban Polymers Private Limited, with revenue of Rs 880 crore. On revenue the target is a little over a quarter the size of the buyer, at 27.7 per cent. A target that size gets read properly rather than glanced at.

Start with the fact that shapes every question asked inside this room. Sundarban Polymers Private Limited earns EBITDA of Rs 132 crore on revenue of Rs 880 crore, a margin of 15.0 per cent. Harivansh Packaging Limited earns exactly the same 15.0 per cent margin on its own Rs 3,180 crore. Because the two businesses are equally profitable on revenue, nothing found in this room can be explained as a margin story. The diligence questions are forced towards price, structure and funding instead. That is worth sitting with. The buyer cannot tell itself that it is buying an unusually profitable business, and it cannot tell itself it is fixing an unprofitable one. Whatever this transaction turns on, it is not the margin.

Now the staging. Stage one carries what supports the shape already agreed at the term sheet: the trading record behind the Rs 132 crore, the borrowings behind Sundarban Polymers Private Limited's net debt of Rs 180 crore, and the working capital position that will later be measured against the Rs 96 crore peg written into the agreement. Each of those is there because a buyer confirming an offer needs it. None of them is there because a buyer signing an agreement needs it.

Stage two carries the second set. The contracts with the two counterparties whose consents become conditions to completion sit here. A buyer cannot sensibly accept those consents as conditions without having read what the underlying contracts say. So do the property and employment records the agreement will have to describe, and anything else a condition turns on. Stage two opens as documentation begins, the milestone after confirmatory diligence.

Then the restriction. Because the two businesses sell to some of the same customers, customer level pricing does not sit in either open stage. The pricing goes to a restricted group who read it and report conclusions, and the transaction team under Ashwin Rege receives those conclusions rather than the file. The team ends up knowing whether the pricing behind the Rs 132 crore is stable. The team does not end up knowing what any named customer pays.

The record carries no document count for this room, no number of diligence questions asked, and no list of findings. A room holding around 6,000 documents, with 400 questions raised against it, would sound authoritative and would settle nothing. The structure of the room is what matters, and the size of a room says nothing about the quality of the disclosure in any case.

How does a transaction team actually use the room?

Four people look at the same room and see four different objects, and that is the most practical way to hold everything above.

Ashwin Rege, running the buyer's transaction team, treats the room as a list of things that could stop the deal or move the price. His team works from its own question list, not from the folder tree, and pushes anything unanswered into the written channel. His measure of a good week is not documents read; it is questions closed. He also watches which of his own people have access to what. He will be asked afterwards.

Devyani Kulkarni, as chief financial officer of Harivansh Packaging Limited, is reading for two things at once: whether the trading record supports what was agreed at the term sheet, and whether anything in the room changes the funding conversation she is running in parallel. She is also the person who approves the access list, and therefore the person who has to say no to a reasonable request from a colleague.

The lender underwriting the new borrowing reads the room as a credit file. The lender has no interest in most of it. The lender wants the borrowings, the working capital, the cash conversion and whatever bears on whether this business services debt. Exactly those folders are granted, and the second row of the access grid is sparse for that reason. The lender will do its own reading rather than accepting the buyer's summary.

And the seller's side reads the room as an exposure. Every folder that opens is a folder that cannot be closed again. The seller's people are watching the question log for where the buyer is going, deciding what stage two really has to contain, and keeping the restricted arrangement intact so that the customer pricing that took years to build does not walk out through a transaction that might not happen.

The household version, one last time. The reflex is the same. When a household sells a plot of land, somebody has to hand over the papers. The sensible person in that household makes a list of what was given, to whom, on what date, and keeps a copy of everything handed across. Not because anyone is expected to behave badly, but because six months later a question will be asked and memory will not answer it. A data room is that habit, industrialised, for a transaction where the papers run to thousands and the people run to dozens.