Go-to-Market · August 14, 2026
Your best prompt is still a better question
Why curiosity will outlast prompt engineering.
Prompt libraries multiply like umbrellas in an office: always available, rarely the reason anything got done. A clever prompt can polish an answer. It cannot invent a question worth asking.
Your best prompt is still a better question. Curiosity outlasts prompt engineering because models amplify whatever you aim them at. Vague aims produce fluent mush. Sharp questions produce useful drafts, research angles, and decisions.
Prompt tricks vs good questions
| Task | Bad (prompt theatre) | Good (question first) |
|---|---|---|
| Research | “Summarise this market” | “Who already tried to solve X and why did they quit?” |
| Positioning | “Write a value prop” | “What progress is the buyer hiring us for?” |
| Content | “Make it viral” | “What tension did we see that competitors ignore?” |
| Product | “Suggest features” | “Where do users get stuck before value?” |
**Bad SaaS habit:** prompting for a homepage rewrite without customer language. **Good SaaS habit:** collecting five stuck-moment quotes, then asking the model to draft from those quotes only.
**Bad travel / hospitality habit:** “write Instagram captions about luxury.” **Good habit:** “what anxiety do first-time visitors have about this neighbourhood, and how do we remove it?” Same tools. Different aim.
How to work question-first with AI
- Write the decision you need before you open a tool.
- Turn it into one sharp question a human could answer in an interview.
- Gather notes, quotes, or constraints. Feed those in, not a blank vibe.
- Use AI to expand, cluster, or draft. Keep the question ownership human.
- Reject answers that are fluent but do not change a decision.
- Keep a running list of questions that produced useful work. Retire prompt trivia.
Related: AI can write the post; it cannot be you. For research discipline, First Round Review and Harvard Business Review still reward better questions over better templates.
Welcome to the journal. nau studio starts with research questions so strategy and content have somewhere true to aim.
Visit nau studioContinue reading
If this was useful, read this next
A complementary essay from the same thread.
Swipe left for next, right for previous.
Up next · Go-to-Market
Let AI do the boring bits. You keep the taste.
Scale the busywork. Protect the decisions that define the work.
Read essay →
