Test the strongest reason

The move Give AI the full situation. Then test the single strongest reason, its best counterpoint and what evidence would change the conclusion. If the strongest reason falls apart, the others only make the email longer.

The short version: AI can give you ten polished, plausible reasons without testing any of them. A long list can make a recommendation feel well-supported before you know whether its logic is sound. Testing the strongest reason shows you where the argument is weak.

You're writing an email recommending that a client move more of its Q4 campaign budget to digital. You ask AI for reasons the client should approve the change. It gives you ten polished, plausible reasons, and you choose four to paste into the email.

But you have not tested the logic behind any of them. The number of reasons made the recommendation feel well-supported before you knew whether it was.

Ten reasons might look and feel great but one tested reason is where you find out if it really holds up.

Before you choose which reasons to use, give AI the full situation and ask it to test the strongest one. A reason is only strong if it fits the outcome, audience, constraints and evidence.

Here's the situation:
- What I'm recommending: [your recommendation]
- What outcome matters: [the result you want]
- Relevant context: [audience, constraints, evidence and tradeoffs]

Using only that context:
1. Give me the single strongest reason for the recommendation.
2. Give me the best counterpoint to that reason.
3. Tell me what evidence would change your conclusion.

If the context isn't enough to judge, ask me what's missing before answering.

Filled in, it looks like this

Here's the situation:
- What I'm recommending: Move the client's Q4 campaign budget to
  digital-first instead of the usual print/digital split.
- What outcome matters: Improve response without losing the ability
  to reallocate after early results.
- Relevant context: Their core audience skews 55+. Last year's print
  response ran 3x digital for that segment. Print spend locks earlier;
  digital can move after early results.

Using only that context:
1. Give me the single strongest reason for the recommendation.
2. Give me the best counterpoint to that reason.
3. Tell me what evidence would change your conclusion.

If the context isn't enough to judge, ask me what's missing before answering.

What the test reveals

Strongest reason: Digital-first lets the client reallocate mid-flight based on early results. Print locks spend before performance is visible, which matters most in Q4 when the holiday mix is hardest to predict.

Best counterpoint: Their core audience skews 55+, and last year's print response rate ran 3x digital for that segment. Shifting budget from the channel that converts best for their actual buyers is a bet on format, not results.

What would change my mind: If digital response rates for the 55+ segment have closed the gap since last year, the flexibility argument holds. If the gap is still 3x, print-first with a digital test is the safer call.

Now you can see whether the recommendation is strong enough to send. The counterpoint identified a risk that the original list did not address.

Why this helps: You find the weak point in thirty seconds, before the client finds it.

Context makes the answer relevant, not true. Verify the evidence before you send the recommendation. AI can identify the claim that needs checking. You still decide whether the evidence supports it.