# Decision Journal Entry With Criteria Locked Before You Choose

> Record a big decision the way good forecasters do: weights fixed up front, scored options, probability-tagged predictions and a review date to learn from.

- **Author:** [Elif Yilmaz (@elif_yilmaz)](https://promptabide.com/elif_yilmaz)
- **Tested on:** Claude · Opus 5.5
- **You fill in:** `decision`, `options`, `criteria`, `context`
- **Published:** 2026-08-24
- **Updated:** 2026-09-24
- **Tags:** `decision-making`, `career`, `personal-growth`
- **Keywords:** decision journal template prompt, weighted decision matrix with ai, should i take the startup job offer, pre commit criteria before making a decision, ai prompt to make a big decision
- **Views:** 2401
- **Likes:** 40

**Best for:** Anyone facing a job move, hire, purchase or strategy call who wants a written record they can check their judgment against later.

## Prompt

```
Help me write a decision journal entry. I've already set my criteria and weights below. Do not change them. You may flag a weight that seems to contradict what I've said about my situation.

Decision: {{decision}}
Options: {{options}}
My criteria and weights (sum to 100), set before scoring: {{criteria}}
Facts and constraints: {{context}}

If a critical fact is missing, state the assumption you used instead of asking.

Output:
1. **The decision in one sentence**, with the deadline.
2. **Weight check**: any weight that seems inconsistent with the facts, and why. Keep the weights as given.
3. **Scoring table**: each option scored 1-5 per criterion, a one-line reason per score, weighted totals. Show the arithmetic.
4. **Sensitivity**: the smallest reasonable weight change that would flip the result.
5. **Predictions**: 3 things I expect to be true in 6 months if I choose the leading option, each with a probability.
6. **Tripwires**: 2-3 concrete signals that would mean I should reverse or revisit.
7. **Review date**, and the one question to answer then.

Do not tell me what to choose. The scores are the input; the choice is mine.
```

## Variables

- `{{decision}}` — The choice to be made and its deadline. (e.g. Accept a startup offer or stay; answer due 2 Oct.)
- `{{options}}` — The realistic options, usually 2-3. (e.g. A) Stay. B) Join the startup.)
- `{{criteria}}` — What matters, with weights summing to 100, written before you think about scores. (e.g. Growth 30, Financial security 25, Team 15, Energy/life 15, Optionality 15)
- `{{context}}` — Key facts: money, timelines, risks, what you know and don't. (e.g. Current: $118k, stable, stale stack. Offer: $132k + 0.15%...)

## Example input

```
decision: Whether to leave my permanent senior developer job for a Series A startup. Answer due Friday 2 Oct 2026.

options: A) Stay at my current company. B) Join the startup.

criteria: Growth/learning 30, Financial security 25, Team and people 15, Energy and life outside work 15, Long-term optionality 15.

context: Current job: $118k, 4 years in, great team, stable, but the stack is legacy and I haven't shipped anything new in a year. Offer: $132k + 0.15% equity, AI agent tooling product, 18 months of runway, 14 people, I've met 3 of them. Founders expect "some" weekend work at launch. I have 7 months of savings, no kids, a mortgage.
```

## Output

*Produced by Claude · Opus 5.5.*

```
**1. Decision**
Stay in my current senior developer role or join a 14-person Series A startup. Answer due Fri 2 Oct 2026.

**2. Weight check**
- **Financial security at 25** is reasonable given the mortgage. The 7 months of savings cushions it, though, which may be why it isn't higher.
- **Energy and life at 15** looks low if "some weekend work" turns out to mean most weekends. Keeping it as given.

Assumption: the equity is too small and uncertain to count as financial security. I scored it under optionality instead.

**3. Scoring**

| Criterion (weight) | A: Stay | B: Startup |
|---|---|---|
| Growth (30) | 2: no new shipping in a year | 5: new product, current tech |
| Financial security (25) | 4: stable, lower pay | 3: +$14k, but 18-month runway |
| Team (15) | 5: known and great | 3: met 3 of 14 |
| Energy and life (15) | 4: predictable | 2: launch weekends expected |
| Optionality (15) | 2: legacy stack narrows the next move | 4: agent tooling experience, small equity upside |

- **A:** 2×30 + 4×25 + 5×15 + 4×15 + 2×15 = 60 + 100 + 75 + 60 + 30 = **325**
- **B:** 5×30 + 3×25 + 3×15 + 2×15 + 4×15 = 150 + 75 + 45 + 30 + 60 = **360**

**4. Sensitivity**
Move 10 points from Growth to Energy and life (Growth 20, Energy 25):
A = 40 + 100 + 75 + 100 + 30 = **345**
B = 100 + 75 + 45 + 50 + 60 = **330**
The result flips. The decision hinges on how much you value growth against evenings and weekends. The table can't settle that for you.

**5. Predictions (if B)**
- By April 2027 I'll have shipped at least one feature I'd put on my CV: **80%**.
- I'll be working more than 50 hours in a typical week: **55%**.
- The company will have raised again or be clearly on track to: **50%**.

**6. Tripwires**
- Weekend work in more than 2 of any 4 weeks after launch.
- No fundraising progress by month 10 (8 months of runway left).
- By month 4, I'm still not trusted with meaningful product decisions.

**7. Review date**
Friday 2 April 2027. Question: "Knowing what I know now, would I make the same call, and which prediction was I most wrong about?"
```

## Why it works

The core move is **pre-commitment**: you write weights before seeing scores, and the prompt forbids the model from changing them. That blocks the most common self-deception, which is quietly re-weighting criteria until your preferred option wins.

The **weight check** still lets it notice contradictions without overriding you. **Sensitivity analysis** shows how fragile the result is. If a 10-point shift flips it, the table isn't really deciding anything and your gut is. **Probability-tagged predictions** plus a **review date** make the entry something you can later grade, which is the whole point of a decision journal: over time you learn whether your judgment is calibrated. "Do not tell me what to choose" keeps ownership where it belongs.

## When not to use it

Overkill for reversible, low-stakes choices; just pick. Weighted scoring also flattens decisions that hinge on a single deal-breaker (a visa, a health need, a partner's job); name that constraint first and only score options that survive it. Don't use it for legal, medical or tax decisions in place of a professional.

---

Canonical HTML: https://promptabide.com/bides/decision-journal-entry-with-precommitted-criteria
Agent guide: https://promptabide.com/llms.txt · https://promptabide.com/agent-instructions.md
Sitemap: https://promptabide.com/sitemap.xml