Lesson 11 of 25

The Art of the Open Question

A question that can be answered yes or no will be answered yes or no. If instead you name the dimensions you want the answer to cover - why, how, in what order, measured against what - the model has to actually do the thinking you were hoping for. The work of opening a question is not adding words; it is deciding in advance what a useful answer would contain.

Reach for it whenWhen you get a confident one-paragraph verdict and realise you cannot act on it, because what you needed was the reasoning, the alternatives and the next step.

Open the question by naming the decision you are trying to make and what you will do with the answer, then list the dimensions you want covered as numbered sections. Do not add "think step by step" - current reasoning models already do, and saying it crowds out the instruction that matters. Put the standing background (your business, your city, your budget) in Custom Instructions or memory so each new open question can be short and still land in context.

Worth knowingAn open question invites length, and this platform will fill the space you give it - you can get four screens of strategy you will never read. Bound the answer as you open it: say how long, and say which part you want to be longest. Also ask it to name its assumptions, because an open question gets answered against assumptions you did not state.
  • Name the decision the answer must help you make.
  • List the dimensions as numbered sections.
  • Skip 'think step by step' - reasoning models do it already.
  • Put standing background in Custom Instructions, not in every prompt.
  • Bound the length, or you will get four screens.
Example 1: The Closed vs. The Open Door (Business Strategy)

Instead of

Is social media marketing a good choice for my new cafe in Kondapur?

Perplexity version

I am opening an upscale cafe called 'Kondapur Mornings' in Kondapur, Hyderabad, next month. I have to decide where my limited marketing budget goes in the first 90 days. Help me make that decision rather than telling me social media is a good idea.

Cover these, in this order, with headings:

1. **Who I should actually be targeting.** Kondapur is IT-corridor adjacent - young professionals, young families, people working odd hours. Pick the segment with the best ratio of spend to repeat visits and defend the choice. If you think I am wrong to assume young professionals are the priority, say so.
2. **Platform choice, with reasons.** Instagram and Facebook are the obvious candidates. For each, say what it is actually good for in a neighbourhood food business - discovery, reminder, social proof, bookings - and what it is poor at. Include Google Business Profile and WhatsApp in the comparison, because for a neighbourhood cafe those often outperform the obvious two and most advice ignores them.
3. **A first-month content plan.** Week by week. Concrete post ideas, not categories - "the 6am bean roast, shot on a phone, no caption" rather than "behind-the-scenes content". Say which posts are there to be saved and shared, and which are just there to keep the grid alive.
4. **What I measure, and what I ignore.** Give me three numbers worth watching in month one and name two vanity metrics I will be tempted by. For each real metric, say what number would mean it is working and what number would mean I should stop.
5. **The case against.** One honest section on what organic social will not do for a new cafe, and what else that budget could buy - launch event, delivery platform placement, local print, signage.

Keep the whole thing under 900 words; sections 3 and 4 should be the longest. End with "Assumptions I made" - anything you had to guess about my budget, capacity or competition, since I did not tell you.
Open Perplexity 1,930 characters
Example 2: From a Simple Fact to Deep Understanding

Instead of

Was the Vijayanagara Empire important?

Perplexity version

I want to understand the Vijayanagara Empire properly, not get a verdict on whether it mattered. Treat this as a short tutorial.

Structure it as:

1. **Significance in context.** What changed in South Indian history because Vijayanagara existed - politically, economically, in trade, in religious patronage. Be specific about what it consolidated and what it interrupted.
2. **Architecture: Vijayanagara against Hoysala.** Compare the two directly on materials, plan, scale, ornament and function. Use named examples on both sides - Hampi against Belur and Halebidu - and explain what the differences tell us about the states that built them, since that is the actual point of comparing architecture.
3. **Talikota, 1565.** Why this battle is treated as a turning point rather than one defeat among many. Distinguish between what the battle itself destroyed and what was already weakening - that distinction is where the historiography sits.
4. **Where historians disagree.** Name at least two points of genuine scholarly dispute - the empire's religious character, or the degree to which 1565 was a sudden collapse. Say what each side argues.

Where a date, figure or attribution is contested, say so inline rather than picking one and sounding certain. Finish with three things I should read, and note that you may have the titles slightly wrong - I will check them before buying anything.
Open Perplexity 1,392 characters
Example 3: Seeking a Spark vs. a Bonfire of Creativity

Instead of

Is a historical fiction novel a good idea?

Perplexity version

I want to write a historical fiction novel set during the peak of the Kakatiya dynasty in Telangana - roughly the Rudrama Devi and Prataparudra period. Do not tell me whether it is a good idea. Help me find the book.

Generate five distinct concepts. Distinct means genuinely different in protagonist, scale and register - not five variations on a court intrigue. For each one give me:

- **Protagonist.** Name, station, and the thing they want that they cannot have. Vary the social level across the five: I want at least one concept whose protagonist has no access to power at all - a tank-builder, a weaver, a temple musician.
- **Central conflict.** The pressure that does not resolve on its own.
- **The 'what if'.** The speculative or structural hinge that makes it a novel rather than a chronicle.
- **The problem with it.** Every concept has one - thin plot, unlikeable lead, too much exposition required. Name it for each, honestly. This is the most useful part for me, so do not soften it.

Then a section on commercial potential: rank the five for today's market, separating the Indian English-language market from international, because they want different books. Say what the ranking rests on.

Finally: name the research I would have to do before writing any of them, and flag which of the five depends most heavily on historical detail that may be thin or contested. Do not invent specific chronicles or inscriptions as sources - if you are not confident a source exists, describe the kind of source I should look for instead.
Open Perplexity 1,541 characters