Untangle Any Stuck Decision in 20 Minutes: The Complexity Reducer
When a decision has grown too many moving parts to see clearly, this recipe strips it down to the 2–3 variables that actually drive the outcome, and hands AI the sorting work.
The stuck moment. You run a $2.4M plumbing company. Your biggest commercial account is 22% of revenue, but it pays in 75 days, ties up your two best crews, and blocks the residential work that pays in 7. You've been "deciding" whether to fire this client for four months. Every time you sit down with it, the question mutates: cash flow, crew utilization, the hiring plan, the relationship, what-if-they-refer-someone. You're not deciding. You're circling.
That's the signal for the Complexity Reducer: the decision is tangled, and the fix is less analysis.
The recipe.
- Dump the tangle. Write down every factor, fear, and what-if in one messy list. No organizing. Ten minutes, done.
- Split signal from noise. Ask of each item: if this changed, would my decision change? The 2–3 that pass are drivers. Everything else is a constraint (fixed, so stop analyzing it) or noise (feelings, sunk costs, unlikely what-ifs).
- Collapse to one question. Restate the decision as a single sentence built only from the drivers, with a number in it. "Should I fire Client X?" becomes "Does this account net more than $18K/month after the crew hours it consumes?"
- Test the frame. If answering that one question resolves the decision, you're done. If not, you cut the wrong variables; go back to step 2.
- Decide, and park the rest. The noise items don't disappear; they become separate, smaller decisions for later.
The AI move. Paste this into Claude or ChatGPT:
You are a decision-simplification analyst for a small-business operator. I'm stuck on a decision that has grown too tangled to see clearly.
Context: [Brain-dump your situation here: the decision, your business, real numbers, and every factor swirling in your head. Messy is fine.]
Task:
1. List every distinct factor you find in my dump.
2. Sort them into three buckets: DRIVERS (the 2–3 variables that would actually change the outcome if they changed), CONSTRAINTS (fixed realities; tell me to stop analyzing these), and NOISE (feelings, sunk costs, low-probability what-ifs).
3. Restate my decision as ONE question, using only the drivers, with a specific number or threshold in it.
4. Name the single piece of data that would answer that question, and how I could get it this week.
Format: Four labeled sections, bullet points, under 400 words.
Constraints: Do NOT recommend a decision. Do not invent factors I didn't mention. Be blunt about what belongs in the noise bucket.
What it's worth: Four months of circling on a 22%-of-revenue decision, compressed into one answerable question. That's the difference between deciding this week and dragging your cash flow for another quarter.