Stop Asking AI for Answers (Ask for Questions Instead)

The 5-step system for generating better AI questions.

Back in August/September last year when I made decision to focus more on writing for AI Maker Labs for my paid subscribers, I had this feeling looming over my head:

“How do I convert more free subscribers to paid?”

Just like any other AI users, I went to Claude and asked the solution.

Claude gave me the playbook:

• Create a compelling lead magnet

• Add social proof and testimonials

• Show a clear value comparison chart

• Build a conversion email sequence

• Offer a limited-time discount

It was thorough. It was actionable.

And I had absolutely no idea which one would actually work.

So I saved it to my notes folder with the other 50+ frameworks I’d never implemented. Why? Not because the solutions were wrong. Because I had no way to choose between them.

Every answer felt useful in the moment. Yet, every answer just made the decision harder.

I thought the problem was me. Maybe that I wasn’t disciplined enough to just pick one and execute.

Turns out the problem was that I was asking for solutions when I didn’t know what problem I was solving.

What Changed Everything

So, I tried asking a different question.

Instead of “How do I convert more free subscribers to paid?”

I asked:

“What questions do I need to answer to convert more free subscribers to paid?”

Here’s what came back:

• Do you know WHY people aren’t converting? Have you actually asked them?

• What percentage of your audience will never pay, regardless of what you offer?

• Are people not converting because they don’t see the value, or because they see it but don’t need it yet?

• Which of your free posts do paid subscribers read before converting? Is there a pattern?

• Are you selling what people want, or what you think they should want?

I stared at that list.

I couldn’t answer a single one.

And that’s when it hit me: Conversion was never the problem. My clarity was.

I was trying to optimize conversion before I knew what was actually blocking it.

But after running through these exercises all over again, here’s what I learned:

Answers give you options. Questions give you the criteria to choose between them.

And when you’re stuck, criteria is way more valuable than options.

Why This Is Hard to See

Here’s why we keep asking for answers instead of questions.

Answers feel productive. Questions feel like procrastination

When AI gives you “10 Ways to Increase Conversions,” you feel like you’re making progress. You’ve got a plan. You can take action.

When AI gives you “10 Questions You Need to Answer About Conversion,” it feels like you’re going backwards. More research? More analysis? You just want to fix the problem.

I get it. I felt the same resistance.

But one thing I realized was acting on the wrong problem is way more expensive than taking time to understand the right problem.

Instead of giving yourself more work, working on the right questions can save you from wasted effort (there's also a deeper reason this matters—passive AI use may actually be shrinking our cognitive capacity).

Stop Reading. Try This Right Now.

Before I explain how this works, let’s try it.

Think of a decision you’re stuck on right now. Not a tactical question—a real strategic decision where you’re not sure what the right move is.

Got it?

Now open Claude (or ChatGPT, whatever) and ask:

“What questions do I need to answer to [your decision]?”

Go ahead. I’ll wait.

.

.

.

Okay, look at what came back.

Notice anything?

Probably half those questions made you uncomfortable. Because they revealed something you don’t actually know yet.

Maybe one of them made you realize you’re trying to solve the wrong problem entirely.

That slight discomfort you’re feeling right now—that’s the insight trying to surface.

What Just Happened

Instead of getting tactics you will never use, you got clarity about what you’re missing.

And here’s the thing: you can’t skip this step.

You can collect all the frameworks you want. You can implement all the tactics. But if you don’t know what problem you’re actually solving, you’re just guessing with confidence.

The questions reveal the gap between where you are and where you need to be to make a good decision.

The AI Question-Generation System

Okay, here’s how to actually do this systematically.

This isn’t about replacing answer-seeking entirely. Sometimes you just need the answer. “What’s the syntax for this Python function?” doesn’t need a Socratic dialogue.

But strategic decisions? When you’re genuinely stuck? That’s when questions beat answers every time.

Here are the techniques I use:

---

1. Start With “What Questions Do I Need to Answer?”

This is the foundation. The one prompt that changes everything.

When you’re stuck, don’t ask AI for the solution. Ask:

“What questions do I need to answer to [your stuck point]?”

Examples:

• “What questions do I need to answer to grow my newsletter?”

• “What questions do I need to answer to choose between these two tools?”

• “What questions do I need to answer to decide my content strategy?”

This reveals the gap between what you know and what you need to know.

And here’s the key: pick the question that makes you most uncomfortable.

That’s usually the one you’ve been avoiding. That’s usually the one that’s blocking you.

---

2. Layer Your Questions (Don’t Stop at the Surface)

Most people ask one question and stop. That’s like doing one rep and calling it a workout.

Try this instead:

First layer: “What questions do I need to answer about newsletter growth?”

AI gives you surface-level questions like:

• What’s my current growth rate?

• Which channels drive the most signups?

• What’s my conversion funnel?

Second layer: “For each of those questions, what questions does that raise?”

Now you get deeper:

• What does “growth” actually mean for my specific goals? (Revenue? Reach? Impact?)

• Which channel drives quality subscribers vs vanity numbers?

• Am I measuring the right conversion points, or just the easy ones?

Third layer: “What would someone who’s already solved this ask about these second-order questions?”

Now you’re getting expert-level insight:

• What would I need to believe about my audience for growth to matter right now?

• Which growth constraint should I solve first—traffic, conversion, or retention?

• Am I optimizing for the business I have or the business I want?

Each layer reveals different insights:

• First layer: Your immediate knowledge gaps

• Second layer: Hidden assumptions and dependencies

• Third layer: Expert-level pattern recognition you can’t see from where you’re standing

---

3. Borrow Other People’s Brains

This one’s sneaky good.

When you’re stuck, you’re stuck in YOUR perspective. Your mental model. Your assumptions.

But AI can inhabit someone else’s brain for you.

Ask it to generate questions from different expert perspectives:

“What questions would a [specific expert] ask about [your problem]?”

Examples I’ve used:

• “What questions would a cognitive psychologist ask about AI adoption?”

• “What questions would a media company executive ask about newsletter growth?”

• “What questions would a business strategist ask about my product decision?”

• “What questions would a skeptic ask about this approach?”

Each perspective reveals questions you’d never generate on your own.

I did this when I was stuck on newsletter strategy. I asked:

“What questions would a media company executive ask about growing a newsletter?”

One question that came back: “What percentage of your audience will never pay, no matter how good your content is?”

That question changed everything. I realized I was optimizing for total subscriber growth when I should’ve been optimizing for best-fit subscribers.

Sometimes having more information is not what you want. Sometimes what you need is simply a different view point. I took this even further by reprogramming my AI to actively disagree with me—different technique, same principle.

---

4. Work Backwards From Success

Here’s the problem with most questions: we start from where we are and ask “what should I do next?”

But we haven’t defined what “success” actually looks like.

So we optimize for motion instead of direction.

This technique flips it. Start from the end state and reverse-engineer what you’d need to know to get there.

Ask AI:

“If I’d already solved [your problem] perfectly, what questions would I have needed to answer along the way?”

This does two things:

1. Forces you to define what “solved” actually means (you’d be surprised how often we skip this)

2. Reveals the knowledge gaps between where you are and where you want to be

Example from when I was stuck on content strategy:

I kept asking: “What should I write about next?”

AI kept giving me topic ideas. I kept not using them.

Then I tried:

“If my newsletter was working perfectly in 6 months—growing steadily, converting well, and readers loved it—what questions would I have needed to answer to get there?”

Here’s what came back:

• What does “working perfectly” mean? Revenue target? Subscriber count? Reader impact?

• Which readers do I want more of? Which readers am I okay losing?

• What would those ideal readers need to see to convert?

• What’s the relationship between growth content and conversion content?

• Which topics do my best subscribers care about vs what attracts tire-kickers?

Here’s the key lesson I learned: All these questions help me clarify what success means, what I’m willing to sacrifice for it, and force me to confront what success actually looks like before I start optimizing tactics.

Most stuck-ness isn’t “I don’t know what to do next.” It’s “I don’t know what I’m trying to build.”

---

5. Turn Questions Into Experiments (Don’t Just Think—Test)

Questions without experiments are just interesting thoughts.

After you generate your questions, ask:

“If I don’t know the answer to [question], what’s the fastest way to find out?”

This turns philosophical questions into something that you can test.

Example from my conversion research:

1. Question: “Are people not converting because they don’t see the value, or because they see it but don’t need it yet?”

2. Experiment: “What if I survey recent free and paid subscribers and ask them directly?”

I did that survey. 160 responses. Turned out:

✅ 39% said they were overwhelmed by too many AI options.

They weren’t converting because they were decision-paralyzed, not because they didn’t see value.

That one data point changed my entire content strategy.

Instead of “here are more AI tools you need to use,” I shifted to “here’s how to simplify your AI stack with more advanced workflow.”

One experiment. One answer. Complete strategy shift.

If you’ve been reading AI Maker for a while, you know I don’t really share “cool new AI tools that will change your life.” I stick to a few tools and go deeper than surface-level content. That’s how I differentiate.

When to Use Questions vs When to Use Answers

Look, I’m not saying you should question everything. That’s just paralysis with better optics.

Here’s when to use each:

Use questions when:

• You’re stuck on the same problem after trying multiple solutions

• You’re about to make a decision you’ll live with for months

• You’ve collected advice you haven’t implemented

• You don’t know why previous tactics failed

Use answers when:

• You’ve already done the research and just need execution tactics

• It’s a straightforward how-to question

• The cost of being wrong is low

• You can iterate quickly if you’re wrong

The skill isn’t using questions OR answers. It’s knowing which mode you need.

Most people default to answer-seeking because it feels faster. But when you’re solving the wrong problem, fast execution is just fast failure.

Note: If you want a practical decision filter for this, I have a full guide on when NOT to use AI at all.

What Happened to My Conversion Problem

Remember those 5 conversion tactics from the beginning?

I stared at that list for weeks. Which one should I build first? All of them sounded good. All of them could work.

Then I asked better questions. I ran that survey. I found out 39% of my readers were overwhelmed by too many AI options.

That changed everything.

The problem wasn’t lack of value. It wasn’t lack of trust. It wasn’t even pricing.

The problem was decision paralysis.

That one insight gave me clarity on what to prioritize.

I didn’t need all 5 tactics. I needed to understand which ones addressed the real constraint.

So I implemented social proof and limited-time discounts first—both reduce decision friction.

I planned the lead magnet for later—once I had more clarity on what simplified approach would resonate most.

The comparison chart and email sequence? Still on the list, but not the priority right now.

The questions unlocked my clarity. They gave me the criteria to know WHICH tactics to implement first, WHAT success looks like, and WHICH metrics actually matter.

Instead of guessing which of 5 options might work, I had a hypothesis I could test.

The questions I asked in September shaped the newsletter you’re reading right now.

Your Turn

You’ve got a decision you’re stuck on. You know the one.

Here’s what I want you to do:

Open your most favorite AI. Don’t ask for the solution.

Ask:

“What questions do I need to answer to [your stuck point]?”

Look at what comes back. Pick the question that makes you most uncomfortable.

That’s your starting point.

You don’t need more answers. You need better questions.

And now you know how to generate them.