Almost every legal AI tool on the market now claims to be 'built for due diligence.' Most demo well. Very few were built by anyone who has actually run a DD process, and the difference only shows up once you are three weeks into a deal with a data room that keeps moving.
These are the nine questions the sharpest partners ask us when they are evaluating Deeligence. Ask them of every vendor on your shortlist, including us.
1. Was the AI built for diligence, or for contracts generally?
Why it matters. A general-purpose tool needs a skilled operator: someone who knows what to ask, how to phrase it, how to spot a wrong answer, and how to iterate. In practice that means pairing a technical person with a lawyer, and most firms cannot sustain that. The gap is not lawyer skill. It is whether the thinking was done calmly in a prompt library months ago, with the edge cases worked through, or at 9pm on a Tuesday under deal pressure.
What a good answer sounds like. The tool already knows what a lawyer is looking for in a share sale, an asset sale, a regulated target. Your team is not writing prompts from scratch or adapting generic templates.
How Deeligence answers it. Our AI fields are built by ex-BigLaw lawyers working alongside machine learning engineers, focused only on diligence and only on the outputs deal lawyers actually need.
2. How accurate is the contract extraction, and at what granularity?
Why it matters. AI contract review covers everything from a one-paragraph summary to a structured, field-by-field extraction a partner can rely on. Those are very different products.
What a good answer sounds like. A specific field list, clarity on whether local law context is baked in, and an output format a lawyer can verify line by line rather than re-read in prose.
How Deeligence answers it. The AI Scanner pulls 1,000+ fields from contracts into a reviewable table. On a recent deal, HopgoodGanim Lawyers measured an 83% increase in contract review speed using Deeligence.
3. What happens when the data room changes?
This is the question tools built by non-practitioners consistently miss. Working out what has been added, replaced or quietly removed across thousands of documents takes hours, sometimes half a day, every single time the seller uploads.
It is also not a probabilistic problem. A document either changed or it did not. A tool that leans on a language model to describe changes in plain English is solving a softer, fuzzier version of the question, and sooner or later it will let a swapped document through because nothing flagged it.
What a good answer sounds like. Deterministic change detection, document by document, with a clear record of what moved and when.
How Deeligence answers it. Change Tracker is purpose-built for this. When the seller updates eight documents on day six, your team is told exactly where to look. On an average deal, that one feature saves 30 to 60 hours.
4. How do I see where the team is up to?
Diligence is a team sport, and hours vanish chasing colleagues for status updates they forgot to send. Ask whether the tool has any real concept of allocation and re-allocation: which senior reviewed a document, which partner signed off on an issue. Ask, too, whether 'using AI' in practice means five associates running five separate chats with no shared view of what has been covered.
And when an issue you raised about an expired contract needs updating because the seller has produced a renewal, who rewrites the summary with the new price and marks the question answered?
What a good answer sounds like. Allocation, status and sign-off live in the platform, with an audit trail, not in a spreadsheet someone maintains by hand.
How Deeligence answers it. Work is allocated in-platform, status is visible to everyone on the deal in real time, and every action is logged.
5. Is it AI contract review, or does it run the whole diligence?
The biggest time sink on a deal usually is not reading contracts. It is stitching everyone's findings into one coherent report. A tool that only accelerates contract review leaves the larger problem untouched.
"You would think the time saved would be the AI, but my favourite part is the project management: figuring out what to review, saving it somewhere, finding a template, working out where things go. That is where Deeligence really helps." Thomas Hill, Senior Associate, Cowell Clarke
What a good answer sounds like. A vendor who can split their claimed time savings between review and everything around it, with a breakdown across a full deal rather than a single document.
6. Does the output land in the firm's own template?
Excellent analysis inside a chat window has not moved you much closer to a client-ready deliverable. Building some form of export is easy. Automating one that holds a firm's house style is considerably harder.
What a good answer sounds like. Ask to see the real export, not a screenshot. Is it a Word document in your house style, ready for partner sign-off? Or an internal transcript that needs hours of reformatting before it can leave the building?
How Deeligence answers it. Reports export straight into your firm's Word template, so what reaches the partner's inbox is ready to review, not ready to rebuild.
7. Which data rooms does it work with, and what happens without download rights?
Is the solution genuinely data room agnostic: does it sit on top of whatever data room the client or the bankers have already chosen? Deeligence works with all of them, and with documents in any form, including zip folders.
What happens when download rights are restricted or a client wants AI switched off? Every AI tool needs access to the underlying files, so no AI review tool can analyse documents it cannot retrieve. There is a persistent myth that a direct data room integration solves this. It does not.
What a good answer sounds like. Honesty about the limitation, plus a clear account of what still works without it.
How Deeligence answers it. We are upfront: without download rights, the AI review cannot run. Everything else does. The dashboard, document assignment, status tracking, change tracker, issues list and report export all still carry the deal.
8. Does it handle Q&A and supplemental questions?
Diligence generates questions, and those questions have their own lifecycle: raised, sent, answered, chased, closed. Managed outside the platform, they fragment across email and spreadsheets.
What a good answer sounds like. A purpose-built Q&A workflow, not a comments field.
How Deeligence answers it. Our Q&A workflow was designed by lawyers to capture the full process, including producing the Excel sheets that load straight into the major data rooms or export for the other side.
9. What happens to material that never lived in the data room?
PPSR, litigation and IP searches do not come from the data room, but they still must land in the diligence report alongside everything else. Ask where they get collated, who reviews them, and how their findings reach the final document.
What a good answer sounds like. Searches are treated as first-class diligence material, allocated and reviewed in the same place as everything else.
How Deeligence answers it. Searches and any other external material are allocated to reviewers and reviewed in-platform, so nothing is lost on the way to the client-ready report. We partner with major providers so searches can be ordered from inside the platform.
Worth knowing which one you are buying
The question is whether you are buying general contract review or a diligence solution. They are not the same purchase, and a platform built to be useful for everything will not have the specificity to carry a deal. Work out which one you need beforehand, not hopes and prayers after the invoice.