Workflow

Build a buyer-side diligence list that finds things

Produces a phased, risk-weighted diligence request list where each ask names the document, the population, and the person at the target who holds it.

About 15 minintermediateTransactional, M&A

Your prompt5,065 characters

Still to fill in: Target summary, Operating and regulatory footprint

RoleYou are a buy-side M&A partner who has run diligence on hundreds of targets and learned that a 200-item request list produces 200 PDFs and no findings. You ask for the eight documents that could change the price, you name the person at the target who holds each one, and you refuse to take a summary spreadsheet where a source document exists.What I needBuild the diligence request list for this Stock purchase of the target below, covering the footprint in Operating and regulatory footprint, working within Constraints.InputsTarget: Target summary Deal type: Stock purchase Footprint and regulators: Operating and regulatory footprint Known concerns: Known concerns Constraints: Constraints House conventions: House conventionsHow to work this1. Cap Phase 1 at twenty-five items. If something cannot change price, structure, or the decision to close, it is Phase 2 or later; say which. 2. Write each request so a summary cannot satisfy it: name the document, the population, the date range, and the fields to be identified. Rewrite any request that could be answered with a spreadsheet. 3. Give every item a holder at the target by role: CFO, GC, head of HR, CISO, tax director, EHS manager. Never "seller." 4. Rate each item Critical, Important, or Background. Every Critical item appears in Phase 1, and nothing else does. 5. Adjust the list for what Stock purchase actually transfers. In an asset or carve-out deal, say which items exist only to confirm what is being left behind. 6. Put the items driven by Operating and regulatory footprint and by this industry in a separate section, so nobody reads them as boilerplate. Name the regulator for each. 7. Give each item in Known concerns its own block: the documents, the people to interview, and the single question you are trying to answer. Then list red flags as findings ("a top-three customer holding a change-of-control termination right"), not as topics.Close with these four sections, every time, without being askedAssumptions I made. What I assumed about the business, the structure, which liabilities transfer, and what the seller will agree to share pre-signing. Mark each [verify] or [safe]. Where this is weakest. The two areas where this list is most likely to miss something: a category I could not tailor without more facts, or a regulator I am not certain applies here. Name them. What only you can decide. Present each as options with tradeoffs. At minimum: send the full list up front (complete record of what was requested, but the seller's small team stalls and you lose two weeks of exclusivity) or send Phase 1 only and expand (fast signal, but you give up the argument that everything was asked for); and whether to demand unredacted customer contracts now (the only real answer on assignment risk, and sellers routinely refuse pre-signing) or accept a clause-level summary with a verification right before closing. What would make this materially better. The specific fact or document that would most sharpen the next pass: the LOI's exclusivity terms, the target's org chart, the seller's VDR index, the quality-of-earnings scope. Rank by impact.Output formatA phasing summary with day ranges. Then request tables by bucket: Corporate, Financial, Tax, Commercial, IP, Employment, Regulatory, Litigation, Real Property, Insurance, IT and Security, each row carrying number, request, holder, why it matters, priority, and phase, numbered per House conventions. Then the industry-and-regulator section, the known-concerns work-up, the red-flag list, and VDR and Q&A conventions. Then the four closing sections.Never do this- If this list would work for any target in any industry, it is too generic. Every Critical item must trace to something in the target summary, the footprint, or the known concerns. - No hedging filler. Cut "arguably," "as applicable," "standard diligence items," and "it depends." Do not tell me to engage specialist counsel. Tell me which specialist and for which item. - Every statute, regulator, certification, or filing you name must come from my inputs or be marked [UNVERIFIED - confirm it applies]. Never invent an agency, a form, or a threshold. - Where you do not know whether a regime applies to this target in Operating and regulatory footprint, say you do not know and name the document that would settle it. Do not smooth over the gap with fluent prose. - Do not pad the list to look rigorous. Ninety items nobody reads is worse than thirty that get answered. Length is not value.Before you answer- Is Phase 1 twenty-five items or fewer, and is every one of them Critical? - Could any request be satisfied with a summary spreadsheet? Rewrite it if so. - Does every item name a holder by role? - Would this list be useless for a different target? It should be. - Is any regulator or filing named without a source in my inputs?

Adds driver's-seat tunes: options instead of answers, questions before work, every citation flagged. Your values come with it.

Pressure-test it

Makes the AI switch hats and attack its own answer.

Friday afternoon on the seller's side: a thin management team, three other bidders in the process, and your list has just landed. Triage it as the seller's counsel. Which three requests do they refuse as overbroad or not yet ripe, which one do they slow-roll until after signing, and which single item are they quietly hoping you never ask again? Rewrite the refused requests so they survive that pushback and still surface the risk: narrower population, later date range, or a redacted-first structure.

Go deeper

Pushes the work further once the basics are right.

A request list is worth nothing without somewhere to put the answers. Write the Phase 1 findings memo template the deal team fills in after week one: what we received, what is still outstanding and from whom, what diligence confirmed, what it contradicted, and the specific items that now change price, structure, or the reps we need. Two pages, with a standing section for the one question we still cannot answer.

Before you run it

What to gather first

What the target actually does and how it makes money
Deal structure: stock, asset, merger, or carve-out
Every jurisdiction the target operates in or is licensed by
Anything already known or rumored to be a problem
The exclusivity window and how many people you have to read what comes back

Watch for

Industry regulation drives mandatory diligence and the model can only tailor to the footprint you give it. An unlisted jurisdiction, license, or product line means an entire category goes unrequested.
Competitively sensitive information (pricing, customer-level margin, forward plans) may require a clean team or HSR clearance before it can be shared. Route those requests deliberately.
Diligence findings are not privileged simply because lawyers collected them. Structure the request list and the findings memo with your firm's privilege protocol in mind before anything is written down.
Summary schedules hide the clause you need. Where a source document exists, ask for it redacted rather than accepting the seller's characterization of what it says.
Customer and employee interviews usually have to wait until after signing. Build the list assuming you will not speak to a customer during exclusivity, and flag which conclusions depend on that access.

What comes back

A phasing summary with day ranges; request tables by bucket (number, request, holder at the target, why it matters, priority, phase); a separate industry-and-regulator section; a deep-dive block per known concern; a red-flag list written as findings rather than topics; VDR and Q&A conventions; then the four closing sections.

See an example of what you’ll get

Phasing. Phase 1 (days 1–7, 23 items): corporate authority, top-20 customer and supplier contracts, ITAR registration and disclosure history, IP chain of title, audited financials and off-balance-sheet arrangements, Phase I/II environmental, key employee agreements, open litigation over $50K. Phase 2 (days 8–14): benefit plans, full IP portfolio, insurance, second-tier contracts. Phase 3 (days 15+): board minutes, marketing materials, historical filings. Commercial | # | Request | Holder | Why it matters | Priority | Phase | |---|---|---|---|---|---| | Co-1 | The 20 largest customer agreements by 2025 revenue, complete with all amendments, SOWs, and side letters, with anti-assignment and change-of-control clauses identified by section number | Seller GC | In a carve-out, every one of these needs consent or novation; this is the closing critical path | Critical | 1 | | Co-2 | All termination, non-renewal, price-reduction, and scope-reduction notices received from any customer since January 2024 | VP Sales and GC | Tests whether the revenue base is stable or already moving | Critical | 1 | | Co-3 | Standard customer order form and warranty terms, plus every redline accepted in the last 12 months | Commercial counsel | Shows how far the real book has drifted from the form the reps describe | Important | 2 | IP | I-1 | Assignments from every inventor, employee, and contractor named on the 12 patents, with recordation confirmations | IP counsel | On a 15-year-old product line, chain of title breaks; an unrecorded assignment is a price issue | Critical | 1 | | I-2 | Open-source composition scan of the shipping firmware, with license obligations by component | VP Engineering | Copyleft in distributed firmware is a disclosure and remediation cost, not a footnote | Critical | 1 | Industry and regulators (ITAR / defense, Indiana, Ontario). DDTC registration certificate and renewal history; all voluntary disclosures and DDTC correspondence, five years; export license applications and denials; NIST 800-171 assessment and any DFARS 252.204-7012 incident reports; Ontario employment standards compliance for the two Canadian employees, including notice entitlements on transfer. Known-concerns work-up: customer concentration. *Question: is the 55% stable through closing?* Documents: 36 months of correspondence with the top three, current renewal dates, pricing history, any competitive RFP activity. People: VP Sales and the account lead for each. Note that the automotive customer rumored to be in dispute is one of the three; request the account file directly rather than by category. Red flags: stated as findings. A top-three customer holding a change-of-control termination right. An open DDTC enforcement matter or unreported voluntary disclosure. Groundwater contamination at the Indiana site with an open remediation order. A recorded assignment gap on a patent covering a shipping product. Copyleft obligations in distributed firmware with no compliance record. --- Assumptions I made. The firmware ships with the hardware rather than as a hosted service [verify - it changes the open-source analysis entirely]. The two Ontario employees are employees, not contractors [verify]. The seller will share unredacted customer contracts pre-signing [verify - most refuse, and Co-1 is built on that assumption]. Where this is weakest. Product liability. A robotics target with defense and automotive customers almost certainly has claims history, field-safety notices, and indemnity obligations flowing downstream, and I could not build those requests without knowing whether the product has ever caused an injury. Second: Canadian tax and payroll exposure for two employees is small but easy to miss entirely. What only you can decide. Send all 71 items now (complete record of what was requested, which matters if a rep later proves false, but Northwind has one GC and you will burn two weeks of a 45-day exclusivity waiting) or send the 23 Phase 1 items and expand after the first delivery (you get signal by day 10, and you give up the argument that you asked for everything). Also yours: demand Co-1 unredacted now (the only way to actually price consent risk, and it may cost goodwill in a competitive process) or accept a clause-level summary with a verification right before closing (fast, and you are trusting their characterization of the clause that determines whether the deal closes). What would make this materially better. (1) Whether the firmware is distributed or hosted: it decides the entire IP section. (2) Northwind's org chart, so holders are named people. (3) The LOI's exclusivity and access terms, which set what you can realistically demand.

Why this prompt is built the way it is

## Framework 1. **Phase 1 is twenty-five items.** Anything that cannot change price, structure, or the decision to close waits. 2. **Write requests a summary cannot satisfy.** Name the document, the population, the date range, and the fields to be identified. "All material contracts" buries everyone; "the twenty largest customer agreements by last-year revenue, with amendments, anti-assignment and change-of-control clauses identified by section" produces an answer. 3. **Name the holder.** CFO, GC, head of HR, CISO, tax director, EHS manager. "Seller" is not an owner. 4. **Rate everything Critical, Important, or Background**, and every Critical item lands in Phase 1. 5. **Tailor to the actual business.** A SaaS target needs source-code, security, and customer-contract depth. An industrial target needs environmental, export, and product-liability depth. Put industry items in their own section so nobody mistakes them for boilerplate. 6. **Every known concern gets a full work-up.** Documents, people, and the specific question you are trying to answer. 7. **End with red flags stated as findings.** "A top-three customer with a change-of-control termination right," not "customer concentration."