AI Prompts for Decision Making

Most decision-making frameworks are designed by people who study decisions in controlled conditions — not by people who have to make them on Tuesday afternoon with incomplete data, three competing priorities, and a boss waiting for an answer. The frameworks are clean. Real decisions are not.

These ai prompts for decision making are built for the messy middle: when you have some information but not all of it, when the stakes are high enough to warrant careful thought but not so high that you can afford to think for a week.

The Problem a COO Faced

Sandra runs operations for a 200-person B2B software company. Her team had to decide whether to migrate their core infrastructure to a new vendor — a decision involving 6-figure costs, 3-month implementation timelines, and significant business continuity risk. She had opinions from 4 different stakeholders, a vendor proposal, and a deadline from her CEO.

The information she needed for a textbook decision didn't exist. Comparable migrations at similar companies weren't public. The vendor's references were, predictably, glowing. And the cost of delay was real — the existing system was showing strain.

The Wrong Approach

Sandra's first instinct was to build a spreadsheet: weight each factor, score each option, add up the numbers. She'd done this before and it usually produced a number that confirmed what she already intuited — a very expensive way to feel validated.

The spreadsheet problem with complex decisions is that the weights are the decision. Once you assign weight 8 to "cost" and weight 4 to "implementation risk," you've already made the decision — the math just launders it. Weighting is judgment, not analysis.

⚠️ Common mistake: Treating structured analysis as objective when the structure itself embeds your assumptions. The best AI decision prompts don't try to make the decision objective — they surface the assumptions so you can examine them.

The Correct Prompts

The Decision Clarifier

Before any analysis, run this:

I need to make a decision about: [describe decision]. Here's my current framing of the choice: [how you're currently thinking about it].

Challenge my framing: (1) Am I solving the right problem? (2) Is this actually a decision or is it a symptom of a deeper choice I haven't named? (3) What would I need to believe for each option to be clearly correct? (4) Am I deciding between options or avoiding a commitment?

What this does: Applies pressure to the framing before you analyze. Many "hard decisions" are hard because they're framed poorly — as false binaries, as tactical choices when they're really strategic ones, or as current choices when they're really about what kind of organization you want to be.

⚡ Pro tip: The "what would I need to believe" output is the most useful. Write down what you'd need to believe for Option A to be obviously correct, then check whether you believe those things. If you do, the analysis confirms an intuition you already hold and can trust. If you don't, you know where the real work is.

The Assumption Excavator

I'm leaning toward [option] in this decision: [describe]. Here are the key assumptions this choice depends on: [list what you think they are].

Now: (1) What assumptions am I missing? (2) Which of my stated assumptions is most likely to be wrong? (3) Which assumption, if wrong, would completely change my answer? (4) What's the fastest way to test the most critical assumption before committing?

What this does: Surfaces the assumption you're treating as fact. Every decision rests on beliefs that aren't verified. This prompt names them, ranks them by fragility, and pushes toward testing before committing.

Results and What Changed

Sandra ran the Decision Clarifier and discovered she was actually making two decisions masquerading as one: "which vendor" and "whether to migrate now at all." The framing had collapsed them together.

The Assumption Excavator revealed that the most critical assumption — "our team can manage the implementation without external support" — was based on one engineer's estimate that hadn't been stress-tested. Verifying it took two days and changed the timeline estimate significantly, which changed the cost-benefit analysis, which changed the vendor recommendation.

The final decision was different from where she started. And it held up — the implementation went smoother than comparable projects precisely because the risks had been named and staffed before the contract was signed.

How to Apply This to Your Situation

For low-stakes decisions you keep deferring:

I've been avoiding deciding about [topic] for [timeframe]. What's likely happening: am I missing information, facing a values conflict, avoiding a commitment, or just suffering from decision fatigue? Based on this description: [context], which is it?

The "which is it" output often resolves the deferral immediately — because the deferral was the decision, just an unconscious one.

For high-stakes decisions with incomplete data:

I need to decide [choice] and I don't have all the information I'd want. Here's what I know: [list]. Here's what I don't know: [list]. What decision would a thoughtful, experienced person make based on the available information, and what would they leave explicit in their documentation so they could revisit it when better information arrives?

⚡ Pro tip: The "leave explicit in documentation" output is the piece most people skip. Good decisions under uncertainty include a stated revisit condition — "we'll check this assumption in 60 days and adjust if X has changed." That's not hedging. It's how you make good decisions in imperfect conditions.

For reversible vs. irreversible decisions:

Is this decision reversible or irreversible? Here's the context: [describe]. If it's reversible: what's the cost of being wrong and correcting course? If it's irreversible: what's the minimum amount of certainty I need before committing, and do I have it?

This prompt distinguishes between two fundamentally different decision types that most frameworks treat identically. Speed is the right optimization for reversible decisions. Certainty is the right optimization for irreversible ones.

Next Steps

Start with the Decision Clarifier the next time you find yourself stuck on a choice that's been sitting on your mental backburner. The three minutes it takes to run the prompt often produces more clarity than the three days of background rumination you've already spent on it.

The Decision Journal Prompt

One insight that most decision-making content skips: the value of recording your decisions before you know the outcomes. Outcome bias — judging the quality of a decision by how it turned out — is one of the most persistent cognitive distortions in professional life. A decision that went well because of lucky external factors wasn't necessarily a good decision. A decision that went poorly for the same reason wasn't necessarily a bad one.

I just made a decision about [topic]. Here's the context, my options, the reasoning that led to my choice, the assumptions I'm making, and the conditions under which I'd revisit this decision: [describe all of these]. Format this as a dated decision log entry I can review in [3/6/12] months.

What this does: Creates a pre-outcome record of your reasoning. When you review it later — after seeing the outcome — you can evaluate the decision process separately from the results. This is how professionals build actual decision-making skill rather than outcome-calibrated confidence.

⚡ Pro tip: Keep a decision journal across all your major choices for 6 months, then run this summary prompt: "Here are 20 decision log entries from the past 6 months: [paste]. What patterns do you notice in: what types of decisions I'm overconfident about, what assumptions I consistently mis-estimate, and one decision-making habit I should change?" This is the self-awareness exercise that separates good decision-makers from people who just happen to have been right.

The Group Decision Prompt

For decisions that require team consensus:

My team needs to decide [choice]. Here are the different perspectives team members hold: [summarize viewpoints]. Design a decision-making process for our next meeting that: surfaces the real disagreement (not the stated one), creates a format for the quietest voices to contribute, and produces a decision everyone can commit to even if they didn't get their first choice.

The "real disagreement vs. stated one" instruction is the key — in most group decisions, people argue about tactics when they actually disagree about values or strategy. Naming the real disagreement shifts the conversation to something solvable. Store the group decision prompt in PromptABCD and use it before any meeting where you need genuine alignment, not just a majority vote.

⚡ Pro tip: Run the Decision Clarifier on decisions that feel obvious, not just hard ones. The highest-risk decisions aren't the ones you agonize over — those get scrutiny by default. The ones you treat as obvious often contain unexamined assumptions that only become visible when you write them down and examine them. Three minutes now beats a lot of regret later. Good decision-making is a skill built through honest retrospection, not just better frameworks.