Six systems your deal could run on. What each one was actually built for.
Every product below is good at the job it was designed for. None of them were designed for a transaction with eleven adversarial parties in it that has to stay provable for seven years. This page says where each one stops, where we stop, and the eight questions to put to any vendor including us.
01 · Origins
A product can only be as neutral as the day-one schema.
Every system carries the shape of the first customer it was sold to. That shape decides who the record belongs to, who can be a participant, and whether the thing survives the close. Features get added. Origins do not get removed.
Origin
The seller’s file cabinet
Built to stage one principal’s documents for controlled release to outsiders. The log is the vendor’s, produced for the party who bought the licence.
Datasite · Ansarada
Origin
Enterprise document governance
Built for regulated file exchange at bank scale — permissions, retention, audit of documents. The deal process itself was never in scope.
Intralinks
Origin
Process management
Built as project software for deal teams — tasks, checklists, pipeline. Sits above the room rather than being it.
DealRoom · Midaxo
Origin
The firm’s CRM
Built to hold one firm’s relationships and pipeline truth. Counterparties are records inside it, not participants with standing.
DealCloud · Affinity
Origin
Nothing — it accreted
Email, spreadsheets, a shared drive and a call schedule. Still the system most deals actually run on, and the one you are really competing with.
Outlook · Excel · SharePoint
Origin
The multi-party record
Built so no participant — including the vendor — is privileged in the schema. Rules are countersigned, actions are chained, custody is delegated to the parties.
Apokto
02 · Teardown
What each one is genuinely good at, and where it stops.
Assessments below are drawn from public product documentation and from bake-offs our team has sat in. If you think one is wrong, tell us on the call and we will change it.
Datasite
Sell-side document staging
Genuinely good at
Bulk ingest and redaction at volume. Bidder-level engagement analytics bankers rely on. Every advisor on your deal has used it before, which is worth more than it sounds.
Where it stops
The room is a publishing surface for one side. Diligence findings, pricing bridges, committee decisions and every post-close consequence live outside it — in email, Excel and your CRM.
On your deal
You will still staff someone to re-key the room’s contents into the IC memo, and the auction’s fairness rests on the seller’s word plus a vendor-produced log.
Intralinks
Enterprise document governance
Genuinely good at
Permissioning and retention at institutional scale, with the compliance paperwork a bank’s second line already accepts. Procurement is frictionless because it cleared review years ago.
Where it stops
Documents are governed; the transaction is not modelled. There is no gate a deal has to pass through, so process discipline stays a human habit rather than a property of the system.
On your deal
Files are safe and findable. The answer to “why did we price it there” is still spread across four inboxes and a deck nobody can locate in eighteen months.
Ansarada
Deal readiness and Q&A
Genuinely good at
Q&A workflow with real routing and escalation — better than most rooms at the thing deal teams complain about most. Readiness scoring gives sellers a genuine pre-process checklist.
Where it stops
Still a seller-side room with a better Q&A module. One answer cannot be released to four bidders in four scoped versions with four independent receipts.
On your deal
Q&A runs more cleanly than it would elsewhere, and stops entirely at signing. Nothing carries into the capital call or the covenant pack.
DealRoom · Midaxo
Deal process management
Genuinely good at
Making an internal team’s workflow legible — playbooks, task ownership, stage discipline across a portfolio of targets. Good value for a serial acquirer running many small deals.
Where it stops
Workflow sits above the documents rather than governing them. Counterparties are guests in your project tool, so there is no custody posture for a CISO to interrogate and no evidence a bidder’s counsel could verify.
On your deal
Your side is organised. The other four parties are still on email, which means the real record is still email.
Email and Excel
The real incumbent
Genuinely good at
Everything, badly, immediately. No procurement, no training, no adoption risk. Every counterparty already has it and nobody has to be persuaded.
Where it stops
There is no record — only correspondence. Nothing is scoped, nothing is provable, and MNPI ends up in the sent folder of whoever left the firm last quarter.
On your deal
It works until the deal is disputed, an LP asks how a number was derived, or a regulator asks who saw what. Then the cost arrives all at once.
03 · The matrix
Ten questions. No checkmark slop.
Where a product does something partially, the cell says what it actually does. Two rows we lose outright, and they are marked as losses.
| Ask every vendor | Apokto | Datasite | Intralinks | Ansarada | DealRoom | |
|---|---|---|---|---|---|---|
| Can every party verify the record without your database? | OFFLINE | NO | NO | NO | NO | NO |
| Are the room’s rules countersigned by all principals, not set by one admin? | 11 PARTIES | ADMIN | ADMIN | ADMIN | ADMIN | NONE |
| Is the vendor architecturally unable to read stored content? | YOUR KEYS | CMK OPT | CMK OPT | NO | NO | NO |
| Does one answer release to four bidders as four scoped versions with four receipts? | FAN-OUT | MANUAL | MANUAL | MANUAL | NONE | NONE |
| Does the record keep running after close — capital calls, covenants, LP packets? | 12 TYPES | ARCHIVE | ARCHIVE | ARCHIVE | PIPELINE | NONE |
| Is the AI scoped by custody lease, with refusals written to the record? | KURT | BOLT-ON | BOLT-ON | BOLT-ON | NO | NO |
| Can a counterparty who refuses to adopt still participate? | BRIDGE | GUEST | GUEST | GUEST | GUEST | NATIVE |
| Is custody delegable up to an air-gapped clean-team environment? | 6 MODES | REGION | REGION | NO | NO | NO |
| LossHave the advisors on my next deal already used it? | RARELY | ALWAYS | ALWAYS | OFTEN | SOMETIMES | ALWAYS |
| LossCan I stand it up for a single asset sale this week, cheaply? | NO | YES | YES | YES | YES | YES |
Assessed against published product documentation as of August 2026 and bake-offs our team participated in. Vendors change; ask them the questions directly rather than taking our word for it.
04 · Where we lose
Four situations where you should buy something else.
A vendor who cannot tell you this is a vendor who has not thought about your deal. These are the cases we walk away from, and what we would tell you to do instead.
Loss 01
You need a room live on Monday for one asset sale
Our implementation is a two-to-four week exercise: custody mode chosen, constitution drafted with counsel, seats mapped to your IdP. If the requirement is staging 4,000 documents for a single divestiture next week, buy a data room. It is the correct tool and it is cheaper.
Come back when the same firm is running its fourth process and re-keying between all of them.
Loss 02
The banker dictates the room and will not move
On sell-side mandates the advisor often specifies the platform, and they have used theirs a hundred times. We do not win that argument on a live deal and we do not try. The email bridge lets you run your side on the record while they stay where they are.
Buy-side and fund operations are where firms adopt us first, because you control those.
Loss 03
You want origination and relationship intelligence
We do not do coverage, contact graphs or business development reporting, and we are not planning to. That is a CRM’s job and the incumbents are good at it.
Keep your CRM. Sync stage and exposure into it from the record.
Loss 04
Your committee needs a long public reference list
The incumbents have twenty years of logos and we do not. If the decision requires naming ten peer firms already live, we will lose it, and saying otherwise would waste your time.
What we offer instead: the architecture in your reviewers’ hands before the first call, and a reference call under NDA.
05 · Due diligence on us
Eight questions to put to every vendor, including us.
Take this into your bake-off. The answers separate architecture from marketing faster than any demo, and a vendor who deflects on questions three and six is telling you something.
See our answers in fullCan we verify the record without your database, after our subscription ends?
Our answer — Yes — verification runs offline against the chained record itself, independent of Apokto’s database or continued existence as a company.
Who sets the room’s rules, and can one party change them unilaterally afterwards?
Our answer — All principals countersign the room’s constitution before it opens, and no single party — including Apokto — can amend it unilaterally afterward.
Name every role at your company that can read our plaintext, and tell us how we would detect it.
Our answer — None, by default: decrypt is lease-gated against your keys, and any break-glass exception is witnessed by every party and written to the record where you can see it.
Show one answer being released to four bidders in four scoped versions.
Our answer — One Q&A answer fans out as four independently scoped versions with four separate, verifiable receipts — ask to see it live in a bake-off.
What happens to the record the day after close, and the day after the fund winds up?
Our answer — It keeps running: capital calls, covenant tracking and LP packets all build from the same close event, and the underlying record stays under WORM retention for seven years.
When your AI is asked something outside its scope, what does it do and where is that written?
Our answer — Kurt refuses and the refusal is written to the record as a scope-wall event — it cannot read across a custody wall it was not leased into, and never silently guesses.
What is the process for a counterparty that refuses to log in at all?
Our answer — The email bridge carries their side of the exchange into the record without requiring them to adopt the platform — full participation is the goal, not a login mandate.
If we are in litigation with you, can we still prove what happened in the room?
Our answer — Yes — the record verifies independently of Apokto’s cooperation, with RFC 3161 third-party timestamping, which is the point of building it this way.
06 · Switching
What it costs to move, stated plainly.
Week 1–2
Custody and constitution
Your security team picks a custody mode. Counsel reviews the constitution template once and it becomes your firm default.
YOUR EFFORT · 6–10 HOURS
Week 2–3
Seats and identity
SAML against Okta, Entra or Ping, SCIM for joiners and leavers. Advisor firms get scoped seats without touching your directory.
YOUR EFFORT · 3–4 HOURS OF IT
Week 3–4
One live process
We start with a real deal, not a pilot sandbox. Prior room contents import as documents; they do not import as a chained record, and we will not pretend otherwise.
HISTORY ARRIVES AS FILES, NOT PROOF
Deal 2 onward
It compounds
The second process starts with the first one’s DRL, redaction sets and issue patterns already loaded. By the fourth, the playbook is the system rather than a document.
SETUP TIME FALLS EACH DEAL