A common question from deal teams evaluating LiquidDocs: "We already pay for Datasite. Why would we need something else?" It is a fair question, and the answer says something about how the industry has come to confuse two different jobs.
A virtual data room is where the documents live. Due diligence is finding out what they say and deciding what to do about it. Ansarada and Datasite are excellent at the first job. Neither was built to do the second, and the gap between the two is where most of a deal's review hours still go.
Strip away the feature lists and a VDR does four things well.
Secure hosting. The seller uploads, the buyer downloads, and nobody outside the permission set sees anything. This is the core product and it works.
Permissions and audit logs. Who opened which document, when, for how long. Sell-side advisors use this to read buyer intent; buy-side counsel uses it to prove what was disclosed.
Q&A workflow. Structured question submission, routing to the right subject-matter owner, and a record of every answer.
Indexing and search. A folder structure, a numbered index, and increasingly some AI-assisted search, redaction, and document summarization on top.
All of this is infrastructure. It makes the review possible. It does not perform the review.
Open a well-organized data room with 1,200 documents. The VDR has done its job: everything is there, indexed, permissioned, and searchable. Now ask the questions the investment committee will ask.
Which customer contracts contain change-of-control provisions, and what share of revenue do they represent? Are there IP assignments missing for any of the engineers listed in the org chart? Do the option grants in the board minutes reconcile to the cap table? Which leases have terms that end inside the earn-out window?
The data room cannot answer any of these. Its search will find the phrase "change of control" in 47 documents. It will not tell you which of those 47 matter, whether the clause is triggered by this transaction structure, or what the aggregate exposure is. That work still belongs to a human, and it is still done the way it was done twenty years ago: open the file, read it, put the finding in a spreadsheet, move to the next one.
Newer AI features in the major platforms narrow this a little. Auto-summaries help a reviewer decide what to open first. Redaction tools save time on the sell side. But a summary is not a finding, and a finding is not a verified finding. Nobody on a deal team is going to put a platform-generated summary into an IC memo without someone qualified having read the source.
It helps to think of the transaction stack as three layers.
Layer 1: the room. Hosting, permissions, Q&A, audit trail. Ansarada, Datasite, Intralinks, and the rest. Mature, commoditizing, necessary.
Layer 2: the review. Reading every document, extracting the terms that matter, cross-referencing across the room, and producing findings a professional will stand behind. Historically this is associates and outside counsel billing by the hour. It is the largest cost and the longest pole in the schedule.
Layer 3: the judgment. Is the exposure material? Does it change price, structure, or the decision to proceed? This stays with the deal principal and their senior advisors, and it always will.
The data room vendors own layer 1. Layer 3 is not for sale. Layer 2 is where the time goes, and it is the layer that has been the hardest to change, because doing it faster without doing it worse requires both machine coverage and human accountability at the same time.
LiquidDocs sits in layer 2 and is designed to work with whatever data room the deal already uses. Documents come out of the room, get structured by AI at full coverage (every contract, not a sample), and each finding that touches price or structure is verified by a qualified analyst before it reaches the deal team. Every finding carries a citation to the source page. Verified findings are marked verified. Unverified ones are not hidden.
The practical differences from relying on the room alone:
You get answers, not search results. A list of every change-of-control clause with the trigger, the counterparty, the revenue attached, and a human's confirmation that the extraction is right.
Coverage is complete. The reason teams sample is that reading is expensive. When reading is done by machine and verification by people, there is no longer an economic reason to leave 60% of the room unopened.
Accountability is explicit. The output is something a lawyer, an accountant, or an analyst has signed off on, with a trail from every conclusion to the paragraph that supports it. That is what makes it usable in a memo.
Yes. This is not an either-or. The room handles disclosure, security, and process. The review handles understanding. Most of our clients run both: the room they already have for hosting and Q&A, and expert-verified review for the findings.
The mistake is assuming that because the documents are in a modern platform, the review is somehow taken care of. It is not. It is just easier to start.
If you want to see what a completed layer-2 review looks like on a real data room, book a call with our team. We can also walk through the side-by-side on our data room platforms comparison page.