CONTEXT(CLONE)
Sign in

Test every claim in the information memorandum against the data room

For: Private equity investment manager or corporate development lead before a binding offer

Pattern: Adversarial reviewNeeds live modelsDesigned for 8 to 400 agents
Get early accessSign in to use

The pain today

The seller's memorandum and management presentation make claims about customers, retention, pipeline, capacity and compliance. The deal team checks the big ones and takes the rest on trust until the price is already set.

The ask

“I attached the information memorandum, the management presentation and the data room documents. Pull out every factual claim the seller makes, look for the documents that support it and the ones that contradict it, and tell me which claims hold, which are contradicted and which have no evidence at all.”

Plain words, as you would say it to a colleague. Edit it to fit your case before you send it.

What you attach or connect

  • Information memorandum and management presentation
  • Data room documents
  • Seller's Q&A answers
  • Financial and commercial fact books

The unit of work

One worker task per one factual claim from the seller's materials.

Why a swarm fits

Seller materials break into many separate claims, each testable against a few data room documents. One side looks for support and another for contradiction, so a claim is kept only if it survives.

Not for

An early look at a teaser with no data room access: there is nothing to test the claims against.

The decision tree

6 typed decisions, each with an action for every answer

At fixed moments in a run, the engine puts one narrow question to a decision model. The decision model never writes text: it answers yes or no with a probability, picks from listed options, or gives a score, about a small slice of the material. The engine then does exactly what this tree says, which is what makes the run auditable. The thresholds are the template's design values, not measured results.

  1. Planner, while planning

    Split or keep serialYes or no, with a probability

    While planning

    Does this sentence of the memorandum bundle several factual claims, such as a retention rate and a customer count, that need separate evidence?

    Sees only: One sentence of the seller's materials

    Why: One claim per unit lets half a sentence be true.

    • Yes: 0.60 or higherthenSplit the work
    • Unsure: 0.40 up to 0.60thenKeep serial
    • No: below 0.40thenKeep serial

  2. Scope checkYes or no, with a probability

    Before work starts on a unit

    Is this a statement of present or past fact that a document could prove, as opposed to a forecast or a description of strategy?

    Sees only: One extracted claim

    Why: Forecasts cannot be tested against a data room and are set aside.

    • Yes: 0.60 or higherthenAccept
    • Unsure: 0.30 up to 0.60thenAccept
    • No: below 0.30thenSkip this unit

  3. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Is the supporting quote from a contract, ledger extract or third-party record, and not from another management-prepared presentation?

    Sees only: The claim, the supporting quote and the source document's type and author

    Why: Management statements cannot prove other management statements.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenMark unresolved
    • No: below 0.50thenReject and retry

  4. Evidence checkA choice among options

    After a worker answers

    Set beside the claim, what does the quoted data room document show?

    Sees only: The claim and the contradicting side's quoted passage

    Why: A contradiction has to be about the same fact to count.

    • Same fact, consistent figurethenAccept
    • Same fact, different figure or datethenMark unresolved
    • Different customer, period or metricthenSkip this unit

  5. Run control, between rounds

    Another round?Yes or no, with a probability

    Between rounds

    Did both sides return no new document for this claim in the last search round?

    Sees only: The documents cited per round for one claim

    Why: Unevidenced claims are reported as such instead of being searched forever.

    • Yes: 0.70 or higherthenStop
    • Unsure: 0.40 up to 0.70thenContinue
    • No: below 0.40thenContinue

  6. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Does this contradicted or unevidenced claim feed a valuation driver such as recurring revenue, retention or capacity?

    Sees only: The claim's outcome and the valuation driver list

    Why: The deal lead decides what reaches the price discussion.

    Accountable: The deal lead decides which contradictions affect price or terms and how to raise them with the seller.

    • Yes: 0.30 or higherthenAsk a person
    • Unsure: 0.10 up to 0.30thenAsk a person
    • No: below 0.10thenAccept

The fleet: who does what

Model tiers by role, not brands: you choose the models. Strong reasoning models plan and reconcile, small fast models do the wide work, and the judge is a decision model from a different family, so it does not share the workers' blind spots.

  1. Planner

    A strong reasoning model extracts each factual claim and names the kinds of document that would prove or disprove it.

    Decisions here:1. Split or keep serial2. Scope check

  2. Workers

    Small fast workers look for support for one claim; workers from another family look for contradiction in the same folders.

    Designed for 8 to 400 agents, one worker task per one factual claim from the seller's materials. Each worker receives only its own unit.

  3. Judge, from a different model family

    A decision model from a third family rules the claim supported, contradicted or unevidenced on the quotes supplied.

    Decisions here:3. Evidence check4. Evidence check

  4. Reconciler

    A strong reasoning model groups outcomes by theme and links contradicted claims to the valuation drivers they affect.

    Decisions here:5. Another round?

  5. Accountable person

    The deal lead decides which contradictions affect price or terms and how to raise them with the seller.

    Decisions here:6. Person decides

Checked before anything is accepted

  • Support and contradiction must both quote data room documents with index numbers
  • Figures in a claim are recomputed in code from the underlying schedules where they exist
  • Management statements are not accepted as evidence for other management statements
  • Claims with no documents either way are reported as unevidenced, not as true

What comes back

  • Claim register: supported, contradicted or unevidenced, with quotes
  • Contradicted claims linked to valuation drivers
  • Questions for management sessions
  • Document requests to close unevidenced claims

What to measure

  • Contradictions the deal team confirms
  • Claims checked compared with the team's manual coverage
  • Post-signing surprises traceable to an unchecked claim

Names of measures only. No result is claimed for this template.

Templates open in the workspace chat with the ask filled in. Nothing runs until you send it.

Get early accessSign in to use

More in Due diligence and M&A

Red-flag customer contracts for change of control and exit rights

For: M&A associate or corporate development manager on a buy-side review

The value of the target sits in its customer contracts, and the data room holds hundreds with order forms and amendments.

Pattern: Map, verify, reduceRuns today6 decisionsDesigned for 4 to 300 agents

Read each disclosure against the warranty it qualifies

For: Transaction counsel for buyer or seller negotiating the disclosure letter

A disclosure letter arrives late with general and specific disclosures and bundles of documents.

Pattern: Cross-examinationNeeds live models6 decisionsDesigned for 6 to 250 agents