The Budget Call: Auditing a Half-Finished Decision Brief
Objective: Given an incomplete internal decision brief proposing a $180K influencer-seeding push, complete the missing expected-value math, classify the reversibility of the call, and flag what the brief omits before it goes to a VP for sign-off in 48 hours.
You're the growth marketing lead at Snowflake reviewing a peer's brief that recommends committing next quarter's discretionary budget to an unproven influencer-seeding channel.
Audit the brief against the lesson's four tools: compute the EV the brief skipped, classify reversibility correctly, and write the pre-mortem nobody wrote.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, shareable, and enough structure to attach directly to the brief for the VP review
The process
3 steps
Step 01 of 03
EV is probability of success times payoff, minus cost. A lower-probability bet with a bigger payoff can beat a safer-feeling one once you actually run the math.
The brief lists three scenarios (conservative, base, aggressive) with estimated probability and pipeline payoff but never multiplies them out against the $180K cost. What's the EV of each, and does the recommended spend clear a reasonable bar?
Procedure
- Enter probability x payoff for each of the 3 scenarios
- Subtract the $180K cost from each to get EV
- Flag whether the brief's stated recommendation matches which scenario actually has the highest EV
Scenario Prob Payoff Cost EV Conservative 60% $220K $180K +$52K Base 35% $500K $180K -$5K Aggressive 15% $1.4M $180K +$30K The brief recommends 'base case', but Conservative has the highest EV once the math is actually run.
Healthy
EV is computed for every scenario before anyone recommends one, and the recommendation matches the highest-EV option or explains why it doesn't.
Unhealthy
A brief names a 'base case' as the recommendation with no EV math behind it, just a gut sense that it sounds reasonable.
What this means
A confident label like 'base case' can hide a scenario that's actually break-even or worse once cost is subtracted.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| The brief's recommended scenario isn't the highest-EV one | Send it back with the EV table attached and ask for a written reason if they still want the base case | 30 min |
Step 02 of 03
Most marketing decisions are two-way doors, walk through and walk back cheaply. The two-door test is what decides how fast and how much rigor a call deserves.
The brief calls this a 'strategic bet requiring exec sign-off,' implying it's a one-way door. The contract is actually a 3-month pilot with a 30-day cancellation clause. Is this really a one-way door?
Procedure
- Check the actual contract terms cited in the brief's appendix, not just its framing language
- Classify the decision as one-way or two-way door based on the real cancellation terms
- Note how the required approval speed and rigor should change if it's reclassified
Contract terms: 3-month pilot, 30-day cancellation clause, no long-term ad spend commitment. Classification: TWO-WAY DOOR (mislabeled in the brief as one-way). Implication: doesn't need a 48-hour VP fire-drill; a 30% pilot budget with a documented off-ramp should be enough to greenlight this week.
Healthy
The reversibility classification is checked against the actual contract terms, not the brief's own framing.
Unhealthy
A genuinely reversible pilot gets treated like a one-way door because the word 'strategic' was in the brief's title.
What this means
Language like 'strategic bet' is often a tell that nobody checked the actual exit terms.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| A two-way-door decision is stuck waiting on a VP calendar slot | Reclassify it and route it through fast-approval instead of the exec review queue | 5 min |
Step 03 of 03
A pre-mortem writes the future failure headline before launch, then asks what assumption would have to be wrong for that headline to be true.
The brief has no pre-mortem section. Write the headline 'This influencer push failed because...' and name the single riskiest assumption it depends on.
Procedure
- Write one specific failure headline, not a vague 'it didn't work'
- Name the assumption in the brief most likely to be wrong (usually the payoff or probability number)
- Propose one cheap way to test that assumption before committing the full $180K
Headline: 'Influencer push failed because seeded accounts had 40% fewer real followers than their stated audience size.' Riskiest assumption: the 'base case' payoff assumes advertised follower counts are accurate. Cheap test: run a $15K seed with 3 accounts first and audit real engagement before committing the remaining $165K.
Healthy
The pre-mortem names a specific, checkable assumption, and proposes a way to test it cheaply before full spend.
Unhealthy
The pre-mortem section, if it exists at all, just restates 'the campaign might underperform' with no specific cause.
What this means
A pre-mortem is only useful if it's specific enough to suggest a cheap test, a vague one is theater too.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| The full budget is about to commit with no assumption ever tested | Carve out a small test tranche tied to the riskiest assumption before releasing the rest | 30 min |
Final deliverable
A completed one-page decision memo with an EV table, a corrected reversibility classification, and a written pre-mortem, ready to attach to the original brief for VP review.
See a reference example
Zendesk, Q1 partnership-spend audit (excerpt) EV table shows the recommended 'high-touch' tier at -$8K EV versus +$41K for the 'self-serve' tier. Reversibility: reclassified from 'one-way' to 'two-way door', 60-day opt-out clause found in section 4 of the vendor contract. Pre-mortem: riskiest assumption is the vendor's stated activation rate, cheap test proposed at 10% of budget first.
Success criteria
You're done when you can:
- EV is computed for all scenarios and compared against the brief's actual recommendation
- Reversibility is checked against real contract terms, not the brief's own framing language
- The pre-mortem names one specific, testable assumption and a cheap way to test it