Diligence risk burn-down

When to use

Use this skill when:

• A company is past first meeting and needs structured diligence
• You need to answer "what could kill this?" quickly
• You need customer, product, GTM, and team evidence for an IC memo

Inputs you should request (only if missing)

• Stage + round dynamics (lead? follow? timeline?)
• A draft "must be true" list from the first meeting
• Access to customer references (including churned / lost deals if possible)
• Data room or key docs (deck, pipeline, financial model, product docs)

Outputs you must produce

1. Ranked risk register (MAX 4 RISKS - force prioritization)
2. Diligence plan (one fastest test per risk, time-boxed)
3. Evidence pack (notes + quotes + data)
4. Anti-confirmation evidence (required: churned customer or lost deal call, competitor validation)
5. Go/No-go recommendation with rationale

Templates:

• assets/risk-register.md
• assets/diligence-plan.md
• assets/customer-call-script.md
• assets/product-eval-checklist.md

Hard rules

Max 4 risks (non-negotiable)

• If you have more than 4 risks, you haven't prioritized.
• The discipline is choosing: what are the 4 things that would actually kill this deal?
• Other concerns go in "watch list" but don't get active diligence time.

One fastest test per risk

• For each risk, define the single fastest test that generates decision-changing evidence.
• "More research" is not a test. A test has a clear pass/fail outcome.
• Examples of fast tests:
  • 1 buyer call with a specific question
  • Pipeline inspection: stage aging, sources, cycle time
  • Hands-on product evaluation with a "blank sheet" re-explain test
  • Churned / lost prospect reference call

Time-box per test

• Every test must have a deadline (hours or days, not weeks).
• If you can't complete the test in the time-box, the risk is either not testable or you need a different test.

Evidence standard

• For each risk, write: "What evidence would convince us this risk is manageable?"
• And: "What evidence would kill the deal?"
• Be specific. "Good retention" is not an evidence standard. "Net retention >110% with cohort data" is.

Procedure

1) Build the ranked risk register (MAX 4)

List risks in these buckets (pick the 4 that matter most):

• Market risk
• Product risk
• GTM risk
• Team risk
• Competitive/moat risk
• Deal/term risk (if relevant)

Rank by: decision impact x uncertainty.

Hard rule: If you can't name the top 4 ways this fails, you aren't diligencing yet.

2) Define the fastest falsification test for each risk

For each of the 4 risks, write:

Risk	Evidence that increases conviction	Evidence that kills	Fastest test	Time-box	Owner
1.				hours/days
2.				hours/days
3.				hours/days
4.				hours/days

3) Run anti-confirmation diligence (REQUIRED)

At least one churned customer or lost deal call:

• If the company has churned customers, you must talk to one.
• If no churn yet, talk to a lost deal (prospect who didn't buy).
• If neither exists, document why and discount retention claims heavily.

At least one competitor/alternative validation:

• Talk to someone using a competitor or the "do nothing" alternative.
• Understand: why didn't they choose this company? What would change their mind?
• If you can't do this, explain why it's impossible and document the gap.

4) Run customer diligence (buyers first)

Do at least:

• 3 buyer calls (people who sign)
• 2 user calls (people who use daily)
• 1 "failed" call (lost deal, churn, or disqualified) - REQUIRED

On calls, capture:

• Trigger event
• Alternatives considered
• What would make them churn
• Expansion path and constraints
• Procurement/security blockers

5) Product diligence (can it survive reality?)

• Request a live demo, then ask the founder to re-explain from scratch.
• Ask for 2-3 hard examples where the product broke and how they fixed it.
• Identify integration points and estimate real deployment cost.

6) GTM diligence (repeatability)

• Validate ICP specificity (not "mid-market enterprises").
• Validate cycle time and onboarding (time-to-value).
• Validate pricing logic and unit economics by segment.

7) Team diligence (learning rate + decision rights)

• Look for:
  • ability to change mind with evidence
  • clear ownership of GTM/product/eng
  • healthy conflict resolution

8) Competitive diligence (why incumbents don't win)

• Identify:
  • direct competitors
  • incumbent "good enough" substitutes
  • internal build threat

Ask: "If an incumbent shipped a V1 in 6 months, what still makes you win?"

Salesforce logging (recommended)

• Ensure an Opportunity exists with a "Diligence" stage.
• Log each reference call as an Activity and tag it (buyer/user/churn/lost-deal).
• Create Tasks for diligence workstreams with owners and due dates.
• Attach/record the risk register + evidence pack (as Notes or Files).

Use salesforce-crm-ops if you need API patterns.

Edge cases

• If the process is rushed: focus only on top 2 risks and explicitly document what you did not validate.
• If references are all friendly: insist on at least one "failed" reference; otherwise discount the signal.
• If you can't get anti-confirmation evidence: document this as a major gap and flag it in the memo.