Questions that get real replies on X, not bait

A question gets replies when answering is easier than scrolling. That means the answer has to be short, the reader has to have one, and the question has to be about something they have actually done or decided. Open questions fail on all three: they are broad, they need a paragraph, and nobody is sure what you want.

The formats below all do the same thing from different angles: they narrow the question until the answer is one line. Options to choose from, a side to pick, a group to speak for, an admission to make. The narrower the question, the more people answer it, which sounds backwards and is not.

• A budget scenario

  Copy

  You have 3 hours a week for marketing a side project. Writing one blog post Replying on X every day One cold email batch Where do the 3 hours go?

  Why it works:

  Three options on their own lines, one constraint, one question. People reply with a word, and the thread becomes a survey everyone can read.

• A scenario with a deadline

  Copy

  Your product has one bug and one missing feature. You can fix one before Monday's demo. The bug affects 5% of users. The feature is the one the prospect asked about. Which one?

  Why it works:

  A real tradeoff with enough detail to argue about. The two sides are evenly balanced on purpose, so the replies split.

• A scenario with a rule

  Copy

  New rule for one month: you cannot post, only reply. Does your account grow faster or slower? I have a guess. Want yours first.

  Why it works:

  The constraint is strange enough to be interesting, and withholding your own guess invites people to commit to theirs.

Name what you do today and ask for one good reason to switch. This beats "A or B?" because you have taken a position, which gives the other side something to push against, and because you asked for a reason, not a vote.

• Tool versus tool, with your current choice

  Copy

  I use a plain spreadsheet for my roadmap, for now. Give me one real reason to move it into a tool. Something that cost you, not a feature list.

  Why it works:

  The "for now" says you are persuadable. Asking for a cost rules out the feature-list replies that make tool threads useless.

• Approach versus approach

  Copy

  Weekly release or ship-when-ready? I switched to weekly in March and the product got calmer. Still not sure it got better.

  Why it works:

  You share your experience and your doubt. Replies come from people with the opposite experience, which is what you want to hear.

• Price versus price

  Copy

  $9 a month or $90 a year, which do you show first on the pricing page? I have tested both and I still do not trust the result.

  Why it works:

  A decision every founder faces, an admission that the data was unclear, and a question with exactly two answers.

Name the group and ask about a moment. The group makes people feel addressed; the moment makes the answer specific. "Founders, how do you stay motivated?" is a group with no moment. "Founders who hired their first person, what did you stop doing?" is both.

• A group defined by a transition

  Copy

  Designers who moved into product management: what did your design training make harder, not easier?

  Why it works:

  The transition defines who should answer, and the twist in the question (harder, not easier) stops people giving the obvious answer.

• A group defined by a decision

  Copy

  People who quit a job to go full time on a side project: what was the number in your bank account the day you did it?

  Why it works:

  A concrete, slightly taboo number. People answer because nobody asks plainly, and the replies are more useful than any article on the topic.

• A group defined by repetition

  Copy

  Anyone who has run a Product Hunt launch twice: what did you do differently the second time?

  Why it works:

  Requiring two attempts filters for experience. The question asks for the delta, which is the most useful part of any experience.

• Be honest, about a habit

  Copy

  be honest: when was the last time you read your own onboarding emails as a new user would?

  Why it works:

  The lowercase opener signals a casual confession, and almost everyone's honest answer is embarrassing, which makes them reply.

• Either or, about motives

  Copy

  Do you build the feature because a customer asked, or because you want to use it? Most of my best features were the second one. Most of my worst were too.

  Why it works:

  Two options and your own answer, which is contradictory enough to make people think about theirs.

• A number as a question

  Copy

  How many hours did you spend on your logo before your first customer? Mine: 14. Not proud.

  Why it works:

  A question with a numeric answer and your own number first. People love to beat or confess a number.

Bait is a question you do not care about the answer to, asked to produce replies. "Coffee or tea?" from a software account is bait. "Which of these two pricing pages can you read faster?" is not, because you will do something with the answer. Readers can tell the difference within a week, and once they have decided your questions are bait, your real ones get ignored too.

Bait replies are also worthless in a practical sense: they come from people who reply to anything, they are not read by anyone, and they teach you nothing about your audience. A question with twelve real answers beats one with two hundred one-word replies, and X's ranking seems to agree, since conversations between real accounts are what keep a post travelling. The engagement guide goes into what that looks like in your own numbers.

1. 01Start with a decision you are making this week. Not a topic, a decision.
2. 02Write the options you are choosing between, each on its own line. Two or three, never more than four.
3. 03Add the constraint that makes it hard: the budget, the deadline, the number of users.
4. 04Decide whether to give your own answer. Give it when you want people to argue; withhold it when you want an honest vote.
5. 05Cut the question down until someone could answer it in a single line while walking.
6. 06Post it in an hour when your audience is around, then stay for the first thirty minutes to reply.

Run out of decisions and you can still find questions in an ordinary day; the prompts for days with nothing to say include several that end in a question. And the fastest way to learn what kind of questions people answer is to answer other people's: the reply guide covers that side.