Deal structure
Asset sale or share sale is not a technical question
Buyers want assets and sellers want shares, for the same reason: one of them leaves the history behind. Where that argument land…
9 November 2026 · 4 min
No single room wins every deal, so this is organised by the process you are running rather than as a ranking. With what each provider actually publishes, rather than what the comparison sites claim.
All M&A Research
Diligence and process
Updated
Key takeaways
The honest answer to which data room is best is that it depends on what you are running. The distance between a two-bidder sell-side process and a forty-bidder auction is wide enough that no single tool wins both, and most of the comparison pages on this subject rank products as though it were.
So this one is organised the way the decision actually gets made. What the room has to do, the criteria worth weighing, what each provider publishes, the shortlist, and then the recommendation by process shape rather than a league table.
One thing is worth knowing before any of it. Every roundup of virtual data rooms carries a pricing table, and almost none of those figures come from the vendors. We went to each provider’s own pricing page in November 2026 and recorded what is actually published. In the M&A tier, hardly anybody publishes a price at all. That is not an oversight, it is a sales model, and knowing it changes how you buy.
Three jobs, and only the first is about storing documents.
It has to disclose the material a buyer needs to confirm what it is buying. It has to demonstrate that the business is organised, because buyers read response speed and completeness as evidence about management and they price it. And it has to evidence what was disclosed and when, which is what decides warranty claims after completion.
That third job is the one sellers underrate. Anything properly disclosed is a known matter: generally excluded from warranty insurance, and the seller’s answer to a claim about something the buyer was shown. A seller who cannot prove what was in the room is relying on the buyer’s recollection of it.
Feature lists are long and mostly identical. These are the things that actually differ, in the order they tend to matter on a sell-side process.
The mechanism behind staged disclosure. You need to set access per group and per folder, not per user and per file, because a process has waves and the waves need different views of the same room.
Test it before anyone is invited. Create a dummy account in each group and look at what it can see. This is the single most common place a room goes wrong, and the failure is disclosing something to the wrong bidder, which cannot be undone.
Below about five bidders, a shared inbox and a spreadsheet cope. Above twenty, the Q&A module becomes the process: questions arriving in volume, routed to the right adviser, prioritised, answered once and published to the right group.
What separates the tiers here is volume handling rather than the existence of a module. Several tools have threaded Q&A. Fewer have bulk question import, priority levels and per-group answer publication.
Two different things that often get discussed as one.
The audit trail is the record: every view, every download, append-only, exportable at completion. That is the evidence layer, and it matters again months later.
Bidder analytics tell you who is actually reading. Page-level dwell time across a bidder set is the closest a seller gets to knowing who is serious before indicative offers land, and on a competitive process that intelligence is worth more than the licence fee.
SOC 2 Type II is the baseline most buyers’ IT functions will ask about. Hosting region matters where the data includes employee records or customer personal data, because the disclosure itself has to be lawful. Check whether the vendor can state where data is held rather than where the company is incorporated.
Measured in how long it takes to get a structured, indexed room from a folder of files, not in how long the signup takes. Bulk upload that preserves your folder structure, and automatic indexing, are the features that save days.
The criterion this page exists to make legible, and the one most comparison sites get wrong. What matters is not the headline number but the basis: flat monthly, per user, per page, or by data volume. A per-page or data-volume contract on a process whose document count you cannot forecast is how a modest deal acquires a five-figure software bill.
Needed on any process involving employee data, customer contracts with named individuals, or a carve-out where some material must not reach a particular bidder. AI-assisted redaction that removes text rather than covering it is meaningfully different from a black box drawn over a PDF, because the latter can often be reversed.
Not the sales engineer. The person who answers at nine at night on the Friday before a bid deadline when a bidder cannot log in. Ask what the support hours actually are and who provides them.
Checked on each vendor’s own pages, 9 November 2026.
| Provider | Price published? | Billing basis | Notes from their own site |
|---|---|---|---|
| Papermark1 | Yes | Flat monthly | Data Rooms from €99/month billed annually (€149 monthly); Plus €249; Premium €549; Unlimited €999. Unlimited documents, rooms and visitors; no per-page or per-GB fees. SOC 2 Type II with EU hosting; open source and self-hostable. 7-day trial of Plus. |
| CapLinked2 | Yes | Flat monthly | Team $399/month with no long-term contract; Enterprise from $550/month or $5,500/year, final pricing custom. 14-day trial. Unlimited guest users. |
| Datasite3 | No | Quote | Its FAQ states pricing is customised for every transaction. The diligence product offers a 90-day free start. |
| Intralinks4 | No | Quote | “Get in touch with sales to get pricing.” |
| Firmex5 | No | Quote | Describes pricing as “always transparent”; every button is Get a quote. Two models: a one-time fee for a single project, priced on data and project length, or an annual subscription priced on annual data volume. |
| Ansarada6 | No | Quote, by data plan | An instant quote builder rather than a price list. Overage is charged on peak usage for the billing period; displayed totals exclude GST. |
Two things follow from that table, and the second is the reason to be sceptical of every other page on this subject.
The per-page numbers you have read are not from the vendors. The range quoted everywhere, roughly $0.40 to $0.85 per page with mid-market M&A rooms “routinely exceeding $50,000”, does not appear on Datasite’s or Intralinks’s own material. It circulates on comparison sites, several of which sell a competing product. It may well be directionally right. It is not sourced, and we are not going to print it as though it were.
Papermark publishes its own comparison carrying figures for competitors. Those are vendor material too, and we have not verified them at source, so they are not reproduced here either.
Quote-based pricing is negotiable pricing. If a provider will not publish a number, the number is a function of what they think you will pay. Get two written quotes, state the data volume and the project length, and ask what the overage rate is before you sign, not after a bidder uploads a 400MB video.
The only entry here that publishes a flat price covering unlimited documents, rooms and visitors, which is why it is the straightforward answer for a small sell-side process.
What it publishes for M&A, from its own product page: bidder groups with per-folder permissions set as full access, view only or no access; a one-click NDA that must be accepted before a document loads, logged against email, time and IP; dynamic watermarking rendering the viewer’s email, IP and timestamp onto each page; an append-only audit log of every view and download; page-by-page analytics showing which bidder read which document and for how long; threaded Q&A routed by workstream and exportable to the closing binder; bulk upload with automatic indexing, or a generated standard index; and at close, the ability to freeze the room, export the audit log and produce an archive matching the disclosure schedule. SOC 2 Type II with EU hosting, open source and self-hostable.
AI redaction sits in the higher tiers, at €549 a month, so a redaction-heavy process should be priced at that level rather than at €99.
Not the answer for: a forty-bidder auction, where the incumbents’ volume handling and bidder scoring earn their price.
The other provider here that publishes a price. Team at $399 a month with no long-term contract, Enterprise from $550 a month or $5,500 a year with final pricing custom, unlimited guest users on both, and a 14-day trial.
Best for: a seller who wants a published, predictable monthly number and does not need the deepest Q&A tooling.
The incumbent most bankers have used, and the one whose process experience shows on a large auction. Its FAQ states plainly that pricing is customised for every transaction. The diligence product offers a 90-day free start, which is a genuinely useful way to build a room before committing.
Best for: large, intermediated, multi-bidder processes where the adviser already works in it.
Quote only, with no published figure anywhere. Its M&A pages advertise the features that matter at scale: bulk question import with priority and access levels, AI-assisted PII redaction, granular permissions, information rights management that retracts a document after download, AI-assisted buyer-interest scoring, and deal tracking.
Best for: the largest processes, and anything where documents may need pulling back after they have been downloaded.
Mid-market, quote only. Describes its pricing as “always transparent” while routing every action to a quote request, which is worth noting. Two models: a one-time fee for a single project, priced on data and project length, or an annual subscription priced on annual data volume.
Best for: buy-side teams doing several deals a year, where the annual subscription beats per-project pricing.
Quote only, built around data plans rather than a price list, with an instant quote builder. Overage is charged on peak usage for the billing period, which is the number to ask about: a single heavy month can reprice the contract.
Best for: sellers who want a structured, guided process and are comfortable with volume-based pricing.
This is the section that should decide it.
Best fit: Papermark, at €99 a month flat. A four-month process at a predictable rate is most of the requirement. You need granular permissions, a watermark, an audit log and staged tiers. You do not need Q&A tooling built for forty bidders, and you should not be buying data-volume pricing for a room that will hold 2GB.
What to avoid: a per-page or data-volume contract for a process whose document count you cannot predict. That is how a £20m deal acquires a £15,000 data room bill.
Best fit: Datasite or Intralinks. This is the category where the incumbents earn their price, and where Papermark is not the answer.
At twenty or more bidders the Q&A module stops being a feature and becomes the process. Several tools here have threaded Q&A; what the incumbents add is volume handling, plus AI-assisted buyer-interest scoring, information rights management that retracts a document after download, and twenty years of bankers who already know the interface.
Best fit: whichever tool has the strongest permissions model and redaction. A carve-out is the hardest disclosure problem in M&A: shared contracts, parent group material, employee data that cannot go out until late, and information that must reach one bidder and not another.
Intralinks advertises AI-assisted PII redaction and rights management that retracts documents after download. Papermark offers AI redaction at its Premium tier. Either works. What matters is testing the group permissions properly before the first bidder arrives.
Best fit: a cheap always-on room. This is not M&A diligence and should not be bought like it. A fundraise needs link-level analytics, per-viewer tracking, an NDA gate and a room that stays open for a year. A flat monthly plan does the job; an enterprise VDR is the wrong instrument.
Best fit: an annual subscription. Firmex’s subscription model and Ansarada’s data plans exist for this: unlimited projects, always-on, priced on annual volume. If you complete four or more transactions a year, per-project pricing is the expensive way round.
Score any provider against your actual deal rather than against a feature grid.
We have not stress-tested these products under load, and this page does not pretend to be a lab review. What it does is record what each vendor publishes, checked against their own pages on a stated date, and set that against the shape of the process you are running.
Where we could not verify something, this page says so rather than filling the gap. That includes the per-page pricing everyone else prints, and it includes any vendor whose current price list we could not locate.
A data room’s job is to show a buyer that the business is organised. That comes from assembling the room against a request list before bidders arrive, putting one person in charge of uploads, keeping a disclosure log, and staging the sensitive material. All of which is in our due diligence checklist, and none of which is a software feature.
We have written separately about what a data room says about a seller before anyone reads it. The short version: buyers read response speed and completeness as evidence about management, and they price it. The tool is the smallest part of that.
Reviewed 9 November 2026. Pricing and features were taken from each vendor’s own published pages on that date and change frequently.
Papermark pricing and its M&A data room page, checked 9 November 2026. ↩
CapLinked pricing, checked 9 November 2026. ↩
Datasite, “How much does Datasite cost?” and its diligence product page, checked 9 November 2026. ↩
Intralinks virtual data room for M&A, checked 9 November 2026. ↩
Firmex pricing, checked 9 November 2026. ↩
Ansarada pricing, checked 9 November 2026. ↩
A secure online repository used to disclose documents to bidders during due diligence, with per-user and per-folder permissions, watermarking, an audit log of every view and download, and usually a question-and-answer workflow. It replaced the physical rooms of printed documents that gave the process its name.
There is no single answer, which is why this page is organised by process rather than as a ranking. A small sell-side process wants a flat monthly fee and strong staged permissions. A forty-bidder auction wants industrial Q&A handling and bidder analytics. A carve-out wants redaction and group permissions above everything else.
It depends entirely on the billing basis. Flat monthly plans are published from roughly €99 to $550 a month. Most established M&A providers publish no price at all and quote per project, priced on data volume or project length, which makes the cost unpredictable for a seller who cannot forecast the document count.
Because quoting per transaction lets them price against the size of the deal rather than the size of the room. It is a sales model, not an oversight. The practical consequence for a buyer of software is that the number is negotiable, and that two written quotes are worth more than any amount of comparison reading.
Dynamic watermarking that renders the viewer's identity onto each page is worth having, because it changes behaviour and because it gives you evidence if a document circulates. Screenshot protection is weaker than it sounds and should not be relied on: anyone determined can photograph a screen. Treat both as deterrents rather than controls.
It routes bidder questions to the right adviser, prioritises them, and publishes each answer to the right group once rather than repeatedly. Below about five bidders a shared inbox copes. Above twenty it becomes the process, and the absence of a proper module is what causes answers to go to the wrong party.
Bidder groups with per-folder permissions, set and then tested before anyone is invited. Create a dummy user in each group and check what it can actually see. The failure mode here is disclosing something to the wrong party, and there is no undo.
For a fundraise or an early-stage information share, sometimes. For a sell-side process, no: you need an audit log you can export at completion, granular permissions and an NDA gate, and free tiers generally omit at least one of those. The entry paid tiers are inexpensive enough that this is a false economy.
Related
Deal structure
Buyers want assets and sellers want shares, for the same reason: one of them leaves the history behind. Where that argument land…
9 November 2026 · 4 min
Exit planning
A trade sale, a sponsor, a management buyout, a serial acquirer or passing it on. Each pays differently, takes a different amoun…
9 November 2026 · 5 min
Valuation
A business with forty per cent of revenue in one account does not get a lower multiple. It gets a lower multiple, a bigger escro…
9 November 2026 · 3 min