A mid-market acquisition generates somewhere between 8,000 and 12,000 documents. Somewhere around document number 6,000, the associate who has been reading contracts since Tuesday starts missing things. That is not a discipline problem. It is a design problem.
I have watched three deal teams run diligence on roughly the same size target in the same quarter. Two of them shipped a clean findings memo inside five weeks. One dragged into month four and nearly lost the buyer. Nothing about the third team was lazier. They just had a worse system for deciding which pages deserved eyeballs.
So here is the workflow that separated them, plus the signal that tells you when it is time to stop managing diligence through shared drives and start shopping for something built for it, which is roughly when you find yourself comparing options like the ones listed on data-room.ca instead of arguing about folder permissions at 11 p.m.
Why diligence slows down
The bottleneck is almost never reading speed. It is decision latency: the hours a reviewer spends unsure whether a document even belongs in scope. Ask a junior associate to assess a 60-page vendor agreement with no framing and they will spend twenty minutes deciding if it is material and forty minutes reading it. Multiply that by four hundred agreements and you have burned a week on triage alone.
Experienced teams kill that latency before it starts. They hand reviewers a scope rule for each folder, written in one sentence, and they make the default obvious: if a contract touches revenue, IP, or headcount, it goes in the review pile. If it does not, it gets tagged and skipped. Reviewers hate this at first because it feels like guessing. By week two they are moving twice as fast and the flagged pile is cleaner.
The three-tier classification
Every document in a deal falls into one of three buckets, and pretending otherwise is where teams lose days. Bucket one is Critical: anything that changes the price or the closing conditions. Purchase agreements, key customer contracts, the cap table, IP assignments, and any litigation file, active or threatened.
Bucket two is Confirmatory: documents that back up a claim you already believe. Employment agreements, standard vendor terms, leases, insurance policies. A reviewer confirms the material terms match the representation and moves on. Ten minutes per document, not sixty.
Bucket three is Context: everything else. Old board decks, marketing plans, internal memos from a reorganization two years back. Context documents matter for understanding the business, but they almost never move the deal. Read them last, or not at all if time is short.
Here is the part most checklists miss. Classify by folder, then audit the classification for a sample of documents in each one. I would rather spend an afternoon spot-checking that a revenue category is right than save the afternoon and find out in week six that half the customer contracts were sitting in Context.
Where contract volume actually comes from
Deal teams underprice one category of document, and it is not the marquee agreement everyone fights over. It is the long tail of routine commercial paper. A typical operating company signs dozens of supplier contracts a year, many of them auto-renewing, many of them carrying change-of-control clauses that nobody in the business ever read.
You do not have to guess whether those clauses matter. The legal framework around what can and cannot be enforced, and how courts read the language, is well documented. According to the Legal Information Institute, the body of contract law that governs how these terms are interpreted is published and searchable, which means your job is rarely to reason it out from scratch. Your job is to find the clause, flag it, and let counsel make the call.
That distinction is worth money. Reviewers who try to resolve legal questions themselves stall out. Reviewers who flag and move on keep the pipeline flowing, because they know legal review is a separate track running in parallel, not a step in their own queue.
A repeatable five-step review loop
This is the loop I would run on any deal under $100 million in enterprise value. It is not glamorous and it works.
- Lock scope before you open a single file. Write one sentence per folder explaining what a reviewer should do with it. Post it where everyone can see it.
- Set a per-document time budget and honor it. Fifteen minutes for Critical, ten for Confirmatory, five for Context. If a document blows the budget, it gets escalated rather than absorbed.
- Log every question, resolve none of them mid-stream. Questions go to one running list with an owner attached. Nothing gets answered by the reviewer who found it, because that reviewer is not the decider.
- Run a daily 15-minute standup on open questions only. No status updates. Just what is unresolved and who is on it.
- Close each folder formally. Once a folder is done, mark it done and stop reopening it unless new information forces it. Reopened folders are where deals quietly die.
Step five is the one teams skip. Someone thinks of one more thing, drags a folder back into the active queue, and now four reviewers are re-reading documents they already cleared. I have seen a single reopened folder add nine days to a timeline. Lock it.
Make the paper trail easy to defend
If the deal ever attracts outside scrutiny, the quality of your process matters as much as the quality of your findings. Regulators and courts care about how records were handled, not just what they said. According to the Federal Trade Commission, how companies maintain and produce business records during investigations is a well-established area of practice, and deal teams benefit from building the file as if someone will read it later.
That means timestamps you can trust, version history that no one can quietly edit, and access logs that show who saw what. It also means a document system that will not let a reviewer overwrite a source file. Spreadsheets and shared folders technically do all of this. They just do it accidentally, and accidentally is a terrible standard when you are being asked to prove you did the work.
The best diligence file is boring. It reads like a checklist, not a novel, and every entry traces back to something a reviewer actually did on a specific day.
When the tool matters more than the process
There is a real threshold here, and teams cross it without noticing. A single deal with a handful of reviewers can run fine on folders and a spreadsheet. The second or third deal back-to-back breaks it. The trigger I look for is when more than five people outside your organization need scoped access to the same set of files, or when you find yourself maintaining an external permission list by hand.
At that point, you are not choosing software. You are choosing how much auditing work you will do when someone asks how the records were controlled. The general pattern that regulators and auditors expect from organizations of any size is documented in the General Services Administration guidance on records and information management, and it is a fair proxy for what a serious deal file should look like.
One more practical note. Do not buy a platform before you know your scope rule. Software amplifies whatever workflow you bring to it, including a bad one. Write the five-step loop first, run one deal on ordinary folders, fix what broke, and only then go shopping. Teams that buy first end up paying for tools they use at twenty percent capacity.
The one thing to remember
Diligence speed comes from decision clarity, not from reading faster. Classify by folder, budget your minutes, log questions instead of answering them mid-stream, and lock folders when they are done. That loop will carry you through deals far bigger than the one in front of you.
So before your next data dump lands, answer one question: does every reviewer on your team know, in a single sentence, what they are supposed to do when they open a folder?
