The Interviewer Prompt: How to Make ChatGPT Interview You Before Doing the Work

You paste a request into ChatGPT, get a generic-sounding wall of text, then spend the next 30 minutes rewriting it so it actually sounds like your business. The problem isn't the model — it's that you handed it half a brief and asked for a finished product.

There's a one-paragraph fix called the Interviewer Prompt, and once you start using it, you'll wonder why you ever wrote prompts the old way. Instead of asking ChatGPT to produce, you ask it to interview you first. It returns five sharp clarifying questions. You answer them in 90 seconds. The output that comes back is dramatically better — not because the model got smarter, but because you stopped guessing at what it needed.

This post gives you the exact prompt, three real use cases (with the actual back-and-forth), and the rules that decide when to use it and when to skip it.

Why first-shot prompts fail

Most solopreneur prompts look something like this:

"Write me a sales email for my new coaching program."

It feels efficient. It is not. ChatGPT now has to silently invent the audience, the offer, the price, the tone, the call to action, the objection it should overcome, and your voice. It will pick the most statistically average answer for every single one of those — which is exactly why the output sounds like every other AI-written email on the internet.

The fix isn't a better prompt. The fix is to stop pretending you can pack all that into a single sentence. Real briefs from human professionals — copywriters, designers, strategists — start with a discovery call, not a deliverable. The Interviewer Prompt makes ChatGPT run that discovery call.

What the Interviewer Prompt actually is

The Interviewer Prompt flips the default flow. Instead of:

Old way: You write a brief → AI produces output → you edit for 30 minutes.

New way: You name the deliverable → AI asks 5 clarifying questions → you answer in 90 seconds → AI produces output that's already 80% there.

It's the same total time investment. But the front-loaded questions force you to make decisions that you'd otherwise punt on, and they pull context out of your head that you wouldn't have thought to volunteer.

It works in any AI tool — ChatGPT, Claude, Gemini, Perplexity. The model doesn't matter. The flow does.

The prompt (copy-paste ready)

Here's the base template. Save it once, reuse it forever:

I want you to help me create [DELIVERABLE]. Before you write anything, ask me 5 clarifying questions that will materially change the quality of the output. Focus on the questions a senior [ROLE] would ask in a discovery call — audience, outcome, voice, constraints, and the one thing I'd regret leaving out. Number them. Wait for my answers before producing anything.

Two slots to fill: the deliverable and the role. That's it.

A few examples of how to fill it in:

• "...help me create a 4-email welcome sequence...questions a senior email copywriter would ask..."
• "...help me create a homepage hero section...questions a senior conversion copywriter would ask..."
• "...help me create a pricing page for my service...questions a senior sales strategist would ask..."
• "...help me create a 90-day content plan...questions a senior content strategist would ask..."
• "...help me create a client onboarding doc...questions a senior client services lead would ask..."

Notice the pattern: the role you ask it to imitate determines the quality of the questions. "Marketing person" gets you generic. "Senior conversion copywriter" gets you sharp.

Three real use cases (with the actual back-and-forth)

Use case 1 — Sales email for a coaching program

Your prompt:

I want you to help me create a 3-email launch sequence for my new productivity coaching program. Before you write anything, ask me 5 clarifying questions that will materially change the quality of the output. Focus on the questions a senior email copywriter would ask in a discovery call — audience, outcome, voice, constraints, and the one thing I'd regret leaving out. Number them. Wait for my answers before producing anything.

ChatGPT's likely questions:

1. Who exactly is this for — what's their role, income range, and the specific pain that drove them here?
2. What transformation does the program deliver in concrete terms (e.g., "10 hours back per week" vs. "feel less overwhelmed")?
3. What's the price, and is there an early-bird, payment plan, or bonus structure?
4. What's the single biggest objection you expect, and what's your honest answer to it?
5. What three words describe how you want to sound — and what three words must you never sound like?

Your 90-second answers are now the brief. The output is night-and-day different from the one-sentence version.

Use case 2 — Homepage hero section

Your prompt:

I want you to help me create a homepage hero section (headline, subhead, CTA) for my done-for-you Instagram management service. Before you write anything, ask me 5 clarifying questions that will materially change the quality of the output. Focus on the questions a senior conversion copywriter would ask in a discovery call — audience, outcome, voice, constraints, and the one thing I'd regret leaving out. Number them. Wait for my answers before producing anything.

Likely questions:

1. Who's the ideal client — solopreneur, founder, or established brand — and what's their monthly content budget?
2. What's the measurable outcome they'll get (follower growth, leads, saved hours)?
3. What do you do that competitors won't or can't?
4. What action should the hero CTA drive — book a call, see pricing, or download a sample?
5. What's the single most common reason a qualified prospect doesn't buy?

You'll notice question 5 in both examples — that "biggest objection" question alone usually rewrites a sales page on its own.

Use case 3 — Service proposal

Your prompt:

I want you to help me draft a 1-page service proposal for a website redesign project. Before you write anything, ask me 5 clarifying questions that will materially change the quality of the output. Focus on the questions a senior agency project lead would ask in a discovery call — scope, outcome, budget, dependencies, and the one thing I'd regret leaving out. Number them. Wait for my answers before producing anything.

Likely questions:

1. Who's the client (industry, size), and what's the business goal driving the redesign?
2. What's the scope — pages, integrations, content, ongoing maintenance?
3. What's the budget range and timeline?
4. What's already decided (CMS, brand, hosting) vs. open for recommendation?
5. What's the success criterion — leads, conversion rate, page speed, brand perception?

This one alone is worth the price of admission. Most solopreneurs write proposals with 30% of the context they need, then wonder why pricing conversations get awkward.

The 5-question rule (and when to break it)

Five questions is the sweet spot. Three is too few — you skip critical decisions. Ten is too many — you lose the speed advantage and the model starts asking redundant variations. If you find yourself wanting more depth, run the Interviewer Prompt twice: once for strategy, then a second pass for execution.

You can also tune the number for the deliverable. A tweet gets 2 questions. A 90-day strategy gets 7. The default of 5 is the right starting point for almost everything in between.

When to skip the Interviewer Prompt

This isn't a universal rule. Skip it when:

• The task is mechanical — "summarize this article in 5 bullets," "convert this to a JSON object," "translate this paragraph."
• You've already done the discovery and just need execution — paste the brief directly.
• The output is throwaway — a brainstorm, a rough draft, a quick lookup.

Use it when the output is client-facing, brand-defining, or decision-driving. Sales emails, landing pages, proposals, content strategy, pricing logic, brand voice docs. Anywhere the difference between "good" and "average" actually shows up in revenue.

Pairing the Interviewer Prompt with other techniques

The Interviewer Prompt isn't a replacement for the rest of your prompting toolkit — it's the front door. Once you've answered the questions, layer in:

• Voice reference. Paste 2–3 examples of writing that already sounds like you and add: "match this voice exactly."
• Constraints stack. "Maximum 80 words. No exclamation marks. No filler phrases like 'in today's fast-paced world.'"
• Output format lock. "Return as a markdown table with columns Subject, Preview Text, Body."
• Self-critique loop. After the first draft: "Now critique this output as a skeptical buyer would. Then rewrite addressing each critique."

The Interviewer Prompt gets you a great brief. These techniques get you a great deliverable.

Why this works (and why you'll forget to use it)

The reason this technique is so effective is the same reason most people skip it: it makes you slow down for 90 seconds before you start. That feels like friction. It isn't — it's compounding leverage. You're investing 90 seconds upfront to save 30 minutes of editing on the back end, and you're producing work that actually sounds like you instead of work that sounds like a default AI assistant.

Save the prompt as a snippet, a saved Claude/ChatGPT custom instruction, or a text-expander shortcut. Once it's a one-keystroke move, you'll start using it for everything that matters.

---

Want a printable cheat sheet of the prompts in this post — plus 7 more solopreneur-grade prompt templates? Download the free AI Morning Brief — our weekly drop of the most useful prompts, tools, and workflows for solopreneurs who actually use AI to run their business.