daBongo LMS AI Training Courses

Claude as a Thinking and Planning Partner

Lesson 4: Comparing Approaches and Stress-Testing Decisions

Log in and enroll to track lesson completion.

Lesson Objectives

By the end of this lesson, students should be able to:

  • Ask Claude to compare multiple approaches using explicit criteria
  • Use pre-mortem prompting to stress-test a decision before committing
  • Apply adversarial prompting to identify the weakest point in their own argument or plan

Lesson Content

Comparison is a strength Claude is underused for.

Claude can hold multiple options in context simultaneously and analyze them against any criteria you specify. This makes it an exceptionally useful tool for evaluating competing approaches, tools, vendors, strategies, or decisions – if you ask for it specifically.

The key is providing both the options AND the criteria. Do not ask Claude to pick the best option without telling it what "best" means for your situation.

Structured comparison prompting.

The most useful comparison format provides:

  1. The options to compare (2-4 works well)
  2. The criteria that matter for your decision
  3. Any constraints that eliminate options outright
  4. What you want Claude to produce: a table, a recommendation, a ranked list, a pros/cons analysis

Example:

Compare these three approaches to onboarding new remote employees: (A) self-paced video library, (B) live cohort training sessions, (C) 1:1 shadowing with a buddy. Evaluate each against these criteria: cost, time-to-productivity, scalability to 50 new hires/year, and manager time required. Format as a comparison table. Do not make a final recommendation – I will decide after seeing the comparison.

Pre-mortem prompting.

A pre-mortem is a technique from project management: before a project begins, imagine it has failed and ask "why did it fail?" This surfaces risks that forward-looking planning misses.

Claude can run a pre-mortem on any plan or decision. Prompt: "Imagine it is [6 months from now] and this [plan/decision/project] has failed. What are the three most likely reasons it failed? For each, what early warning sign would have appeared?"

This prompting pattern produces more actionable risk identification than generic risk analysis because it forces specificity: not "communication might be a problem" but "the team lead left after month two and there was no knowledge transfer plan."

Steelmanning opposing views.

If you are building an argument, planning to make a recommendation, or preparing for a debate or negotiation, ask Claude to steelman the opposing position. "Give me the strongest possible argument against the position I just outlined" or "Argue for Option B as compellingly as you can, even though I'm currently leaning toward Option A."

Steelmanning helps you understand counterarguments at their strongest – so you can address them rather than be surprised by them. It also sometimes reveals that the opposing view has merit you had not considered.

The red team technique.

For plans or proposals that will face scrutiny, ask Claude to play the role of a specific skeptical reviewer: "Act as a risk-averse CFO reviewing this proposal. What objections would you raise and what data would you demand?" The more specific the reviewer role, the more targeted and useful the critique.

Practical Example

An operations manager is choosing between two project management tools.

Generic request:

Which project management tool is better, Asana or Notion?

This produces a generic feature comparison that ignores the specific situation.

Structured comparison:

I manage operations for a 25-person marketing agency. We need a project management tool for client campaign tracking. Our team has varying technical comfort levels. Key requirements: client visibility (clients should be able to view but not edit), task assignment with deadlines, simple status reporting for weekly standup. Budget: under $15/seat/month.
Compare Asana and Notion against these specific requirements. Format: table with our four requirements as rows and the two tools as columns. Note where you are uncertain about current features and I'll verify before deciding.

This produces a decision-relevant comparison rather than a generic feature review.

Lesser-Known Tip

After Claude produces a comparison or recommendation, ask it: "What information would change this recommendation?" This forces Claude to identify the assumptions under which its analysis holds – and surfaces what you should verify before acting on the recommendation. It is far more useful than asking "are you sure?" which produces hedged repetition rather than substantive clarification.

Safety Notes

Claude's comparisons reflect its training data, which has a knowledge cutoff and may not reflect current pricing, features, or availability of tools, vendors, or services. Always verify feature details and pricing directly with vendors before making procurement decisions. Claude's comparison is a structured starting point, not a substitute for vendor evaluation.

Practice Task

Identify a real decision you are currently facing with two or three options. Write a structured comparison prompt with explicit criteria. Ask Claude to produce a comparison table. Then add a pre-mortem: "Imagine I chose Option A and it failed after 90 days. What are the two most likely reasons?" Evaluate whether the pre-mortem revealed anything the comparison table missed.

Completion Check

You should be able to write a structured comparison prompt with explicit criteria, apply pre-mortem prompting to stress-test a decision, and use adversarial prompting to find the weakest links in your own reasoning.

Lesson Quiz

Log in and enroll to take this lesson quiz.

Scroll to Top