Ripping out a data room mid-transaction is the fastest way I know to lose a week of sleep. I’ve watched a CFO do it anyway, three days before a signing, because the export button on his old platform kept timing out. It worked. Barely.
So here’s the deal. Moving platforms isn’t a leap of faith, it’s a procurement project, and you can run it in about two weeks if you know what you’re actually checking. This piece walks you through an evaluation framework I call the Five-Pass Screen, then shows you how to apply it when the incumbent is an enterprise heavyweight and you’re comparing Merrill Datasite competitors against your current setup.
Most people start with pricing. Start with documents instead. Price is easy to read and hard to interpret. Document behavior is where the surprises live, and it’s where you’ll spend your actual working hours.
Why do teams leave an enterprise platform in the first place?
Nobody leaves a tool that’s working. They left because of a mismatch that showed up after the contract was signed.
Three patterns come up constantly:
- Deal size drift. The platform was bought for billion-dollar transactions, but your pipeline now runs on fifty million dollar deals that don’t need the same machinery.
- Pricing opacity. You can’t forecast a bill when every engagement gets a custom quote. Finance teams hate this, and they’re right to.
- Setup time. If your bankers are already working in the room, a two week onboarding window is a nonstarter.
Notice what’s missing from that list. Almost nobody leaves over security architecture. Enterprise platforms are generally strong there. They leave over speed, cost predictability, and whether the tool matches the shape of their actual deals.
The Five-Pass Screen: a framework for vetting any vendor
I built this after running the same evaluation three times in eighteen months and realizing I kept re-learning the same lessons the hard way. Five passes, in this order, because each one can kill a candidate cheaply before you waste time on the next.
Pass one: document behavior
Upload forty real files from your last deal. Not sample documents. The messy ones, with tracked changes, embedded spreadsheets, seventeen versions of the same PDF, a scanned exhibit from 2014.
Then answer three questions. How long until the index finishes? Can you search inside scanned documents? What happens to folder permissions when you drag a file between folders?
That last one trips up more platforms than you’d expect. Permissions that silently reset after a move are a disclosure risk, and you’ll only find it by testing.
Pass two: the exit test
Ask for a full export of a real project, in native format, with metadata intact. Do it during the trial, while you still have leverage.
Vendor lock-in in this category rarely comes from contracts. It comes from the fact that nobody checks how hard it is to leave until they need to leave. Keep your buyer’s checklist honest about this one. Former Federal Trade Commission guidance on consumer and business protections makes the broader point plainly: switching costs and lock-in deserve scrutiny before you sign, not after.
Pass three: security paperwork
Ask for the current report, the date on it, and the name of the firm that issued it. A nineteen-month-old report with a clean opinion is a different document than one issued last quarter.
You also want their encryption standard in writing and a straight answer on where data physically sits. The National Institute of Standards and Technology publishes the framework and standards most serious vendors map their controls against, so knowing the vocabulary helps you ask sharper questions and spot vague answers fast.
Pass four: the pricing model
| Model | What it looks like | Watch for |
|---|---|---|
| Custom enterprise quote | Scoped per deal or per year | Per-page and per-user triggers buried in the order form |
| Flat rate | One number regardless of volume | Storage ceiling that forces an upgrade mid-deal |
| Per-project | Priced by transaction | What counts as a project when a deal restructures |
| Tiered subscription | Monthly bands by seat count | Overage rates for external users |
My take: flat rate wins for anything under a few hundred million in deal value, and it isn’t close. The predictability is worth more than the discount you might negotiate elsewhere.
Pass five: the reference call
Two calls, not one. Ask the first reference what broke. Ask the second how long support took to answer when it did. Vendors hand you happy customers who signed recently. You want the one who’s eighteen months in and slightly annoyed. That’s where the real information lives.
A concrete scenario from a mid-market acquisition
Last spring I sat in on a diligence process for a healthcare services acquisition, roughly eighty million in enterprise value, with maybe six hundred documents in scope. The team ran the Five-Pass Screen over eleven days.
Two candidates failed to pass one. One failed the exit test, badly, which the vendor only admitted when we asked for a live export instead of a description of one. Another looked great until the pricing conversation, where per-page charges pushed the projected bill well past the flat-rate option from a smaller provider.
The winner wasn’t the biggest name at the table. It was the platform that indexed those six hundred messy files in under an hour, exported cleanly, and quoted one number in writing. Six months later the deal closed and nobody thought about the software again, which is the highest compliment a deal platform can receive.
How to run the switch without breaking a live deal
- Pick a gap, not a live period. Between signing and closing is the worst window. Use the quiet stretch after one deal closes and before the next one hits the pipeline.
- Freeze permissions before you migrate. Export the current permission matrix to a spreadsheet first. You’ll rebuild it on the other side, and you’ll want the original to compare against.
- Migrate one deal, not the archive. Old closed deals can wait. Move what’s active, verify it, then move historical projects in batches.
- Run both platforms in parallel for one matter. It costs a little. It buys you a safety net if something’s missing after the export.
- Retire the old account deliberately. Revoke external access, confirm the final export, and keep the archive until your retention window closes. The SEC’s investor resource pages at sec.gov are a reasonable starting point if you need to check what your advisers expect you to retain and for how long.
What the evaluation criteria should never include
Feature count. Every platform in this category has permissions, watermarking, Q&A modules, and audit logs. The differences live in how those features behave under load and how fast a human being can get help when something breaks at 11pm before a bid deadline.
Demo polish matters less than you’d think, too. A slick walkthrough tells you nothing about the seventh week of a nine-month process, which is when diligence fatigue sets in and everybody starts cutting corners on version control.
So run the passes, keep the checklist tight, and give yourself two weeks instead of two days. Your future self, the one staring down a signing deadline, will thank you for it. What’s the one thing you’d refuse to compromise on in your next platform review?
