Vendor demos are rehearsed. They show the platform’s best path through its best features, run by someone who uses it every day. A pilot shows something different: how your own team, with your own documents and no rehearsal, gets on with the software. That is the version you will live with for the length of the deal.
One week is enough if the pilot is planned. Longer pilots tend to drift, and shorter ones skip the outside user’s view, which is where most problems surface. This guide sets out a day-by-day plan and a scoring sheet you can adapt.
What do you need before the pilot starts?
Prepare three things before the account opens: a test folder, a cast of testers and the criteria you will score against.
The test folder. Build a copy of a realistic index with 200 to 500 files. Use real file types and sizes from past deals, but replace sensitive content with dummy documents or redacted copies. Include awkward material on purpose: a scanned PDF, a large spreadsheet with several tabs, a long file name, a folder nested six levels deep and a zip archive.
The testers. You need an admin, one or two internal experts who will answer questions, and two people from outside the deal team playing bidders. Colleagues from another department work well as bidders because they have not seen the platform before.
The criteria. Agree what a pass looks like before anyone logs in. The table further down gives a starting set.
How should the week be scheduled?
| Day | Focus | Who | What to record |
|---|---|---|---|
| Monday | Account setup and bulk upload | Admin | Upload time, files that failed, numbering kept or lost |
| Tuesday | Permission groups and rights | Admin | Time to build groups, inheritance surprises, view-as accuracy |
| Wednesday | Bidder experience | Two mock bidders | Time to first document, search results, viewer speed, mobile use |
| Thursday | Q&A round | Mock bidders, experts, admin | Routing, approval, notifications, answer release |
| Friday | Reporting, export and wrap-up | Admin and deal lead | Activity report quality, archive export contents, support responses |
The order follows a real launch. If Monday’s upload goes badly, you learn it on day one and can ask the vendor for help, which is itself a test of support.
Running a one-week data room software pilot
- Agree pass or fail criteria
List the must-have outcomes and how each will be measured. Share them with the vendor so they can flag anything unrealistic.
- Load the test folder in bulk
Drag the full test index in one action. Record how long it takes, which files fail and whether folder numbering survives.
- Build groups and set rights
Create two bidder groups and one internal group. Set one folder view-only and hide another from one bidder. Check with view-as.
- Invite the mock bidders
Send real invitations to the two outside testers. Ask them to time how long it takes to reach their first document.
- Run a Q&A round
Each mock bidder submits five questions. Route, answer, approve and release them. Note every point where someone was unsure what to do.
- Pull the reports and the archive
Export the activity log, the Q&A log and a full archive. Open the exports and check that they are complete and readable.
- Score against the criteria
Fill in the scoring sheet with the testers on Friday afternoon while details are fresh.
What should the pass or fail criteria be?
Criteria should be specific enough that two people would score them the same way. “Easy to use” is not a criterion; “a new bidder reaches the first document in under three minutes without help” is.
| Area | Criterion | Pass if |
|---|---|---|
| Upload | Full test folder uploads with structure intact | Fewer than 1% of files fail and numbering matches the index |
| Permissions | Rights set per group behave as configured | Both mock bidders see exactly their intended folders |
| Bidder access | Time from invitation to first document | Under five minutes, without calling the admin |
| Search | A phrase inside a scanned PDF is found | The right document appears on the first results page |
| Protection | Watermark on viewed and printed pages | Name and time visible on every page |
| Q&A | One question from submission to release | All four roles completed their step without guidance |
| Reporting | Activity log answers “who opened file X” | Exportable, with user, time and action |
| Archive | Closing export of documents, Q&A and logs | Opens offline and matches the room |
| Support | One question to vendor support | Useful answer within one business day |
Weight the criteria by the deal you expect. A sale with many bidders should weight Q&A and permissions heavily; a single-investor round can weight bidder access and cost more.
What are the most common pilot mistakes?
Most pilots that fail to produce a clear answer fail for the same few reasons.
Only the admin tests. The admin spends the week inside the setup screens and forms a view of the platform from that side. Bidders, who will spend far more hours in the room, never get a say. Always include outside testers.
The test folder is too clean. A tidy folder of 50 PDFs uploads perfectly on every platform. Real rooms contain odd file types, long names and deep nesting, and those are where platforms differ.
Features are tested in a demo account the vendor prepared. A sandbox that the vendor configured hides the setup effort. Start from an empty room so the build time is real.
Support is never contacted. You will need support during the live deal, often late in the day. Send at least one genuine question during the pilot and time the reply.
Which platforms can be piloted without a sales call?
In our facts file, 13 of the 18 platforms list a free trial and eight publish an entry price. The five without a listed trial are all quote-based virtual data rooms, where evaluation usually starts with a demo.
Trial and pricing access across the 18 platforms
A self-serve pilot is usually possible
Cost can be estimated before the pilot ends
Datasite, Drooms, Intralinks, SmartRoom, Venue by DFIN
Run pilots in parallel if you can. Two platforms tested in the same week by the same testers with the same folder produce a fair comparison; two platforms tested a month apart produce memories. If time allows only one pilot, use the software quiz and the feature comparison to narrow the field first, then pilot the top candidate and keep the runner-up as the fallback.
What should happen after the pilot?
Hold a 30-minute review on the Friday or the following Monday. Go through the scoring sheet criterion by criterion and agree a single result for each. Then answer three questions.
First, did any criterion fail outright? A failed must-have usually ends the evaluation unless the vendor can show a fix on your test folder, not on a slide.
Second, what did the mock bidders say? Outside testers notice things the deal team has stopped seeing: confusing folder icons, slow previews, a login step that looks like phishing. Bidder friction has a cost later in the process.
Third, does the commercial offer match what you tested? Confirm that the features you piloted are included in the tier quoted, and that the price covers the room size, users and timeline you expect. The pricing calculator gives indicative totals for comparison.
Delete the pilot room or ask the vendor to delete it, and get confirmation. Even dummy data can contain names and structures you would rather not leave behind.
Frequently asked questions
How long should a data room software pilot last?
Five working days is enough for a structured pilot covering upload, permissions, bidder access, Q&A and reporting. Longer pilots tend to lose focus.
Should we use real deal documents in a pilot?
No. Use a realistic test folder built from dummy or redacted documents with the same file types and sizes. Live deal files should go into the room only after the contract is signed.
Who should play the bidders in a pilot?
Two people who have not used the platform and are not on the deal team. Their first impressions are the closest you will get to a real bidder's experience.
Can we pilot two platforms at once?
Yes, and it gives the fairest comparison. Use the same test folder, testers and criteria for both, and score them in the same session.