Triangulate four sources to find the real wedge in a new customer engagement. Don't trust any single source.
| Source | What it tells you | Failure mode if you rely only on this |
|---|---|---|
| Buyer (CIO / VP / Director) | What they think the problem is | Buyers describe the problem from one altitude up; they often miss the actual failure mode |
| Brief (RFP / written prompt) | What they wrote down to formalize it | Briefs are sanitized; they leave out the political map and the real timeline |
| Industry (benchmarks / peers) | What's actually possible with the technology | Industry benchmarks generalize; your customer's data quality and integration constraints may not match |
| Operator (lead user / daily practitioner) | What's actually broken | Operators are close to the work but may not see the strategic frame |
- List the four sources for the case
- For each, write 1-2 sentences on what they tell you about the problem
- Look for convergence: do all 4 sources point at the same workflow stage?
- If yes, that's the wedge
- If they diverge:
- Operator usually has the truth about where it's broken
- Buyer usually has the budget for what to fix
- You'll need both
- Buyer (Sarah, CIO): "Senior analysts are burning out on the morning-after grind. We lost two in 2025."
- Brief: "Build something that compresses the 4-hour earnings-note drafting time to 30 minutes."
- Industry: Morgan Stanley, BlackRock, and peers have built citation-grounded LLM drafters with HITL approval. Anthropic's Claude is BAA-cleared. the company's deployment patterns apply.
- Operator (Rachel, senior analyst): "Most of my 4 hours is actually verification and formatting. The analysis itself is maybe 45 minutes. If you compress verification, you save 2+ hours."
Convergence: all four point at the verification + formatting layer, not the analysis layer. The wedge is citation-grounded drafting + verification, not "automate the analysis."
When the interviewer asks "how did you decide what to build first", you say:
"I triangulated four sources. Buyer said X, brief said Y, industry pattern said Z, operator said W. They converged on [specific workflow stage]. That's why the wedge is [narrow scope]."
This beats "I'd talk to stakeholders" by 10x in interview signal.
The 4-source convergence test triangulates where the wedge is. The Three Whys (see consulting-frameworks.md) surface the discovery questions you should ask AT each source:
| Why | The diagnostic question | What it surfaces |
|---|---|---|
| System of Record | "If two systems disagree on the data, which one is right?" | Hidden data-trust problems; shadow IT |
| Cost of Inaction | "If we walked away today, what's the cost of this problem persisting for another year?" | Budget defense + sequencing logic |
| Day 2 Ownership | "Who runs this system 90 days after I leave?" | Post-handoff orphaning risk |
Deploy 4-source convergence to find the wedge; deploy the Three Whys to surface the non-obvious risks in week 1.
If sources diverge, the rule is:
- Operator > Buyer on "what's broken" — operators see the daily failure modes
- Buyer > Operator on "what's worth fixing" — buyers control budget and strategic priority
- Brief sets the formal scope — don't violate it without surfacing the violation
- Industry tells you what's feasible — if industry hasn't done it, you probably can't either in v1
The senior FDE move when sources diverge: document the divergence in the field memo. Don't argue with the buyer in week 1; bring the data to week 2.
DISCOVERY (week 1):
Source 1 — Buyer: ____________________
Source 2 — Brief: ____________________
Source 3 — Industry: __________________
Source 4 — Operator: __________________
Do all 4 converge on the same workflow stage? Y/N
If yes → that's the wedge
If no → operator has truth, buyer has budget; you need both