The short version
The best build in public ideas come from what happened this week: something you shipped, a number that moved, a mistake, a decision, a customer conversation, a milestone, or a question for other founders. If you want replies, give the reader a job, like voting or answering in one word. If you want trust or customers, a detailed update can matter more than a fast reply.
Most lists of build in public ideas tell you to share your story. That is a category, not an idea. Below are 30 specific posts, sorted by what you have to share this week. Each has a short example and my rating of how many replies that shape tends to get. The evidence behind the ratings comes after the list.
What should you share when you build in public?
Share what changed this week and give readers a way to join in. A post about your work is only worth following if it has a real detail (a number, a quote, a choice) and a reason to reply.
Start from your week, not from a topic. Every week fits at least one of eight situations: you shipped, you did not ship, a number moved, you made a mistake, you made a decision, you talked to a customer, you hit a milestone, or you just have a question. Find yours in the list below and pick an idea from that group.
This page is about what to post. Which post formats win for founders building in public is covered in how to build in public on X. If you wonder whether it pays off at all, building in public brought me opportunities I no longer needed is the honest version.
How to read the ratings: the reply rating is my own call on each broad shape. I based it on which shapes were most common among my best posts. I did not measure it for these 30 example posts. “High” marks shapes that ask for a short answer (votes, word limits, connect calls and specific questions), plus first-time milestones people want to congratulate. “Medium” marks updates and stories with a real number. “Low” marks advice and cost breakdowns. Low on replies does not mean low value. Those posts often earn saves, trust and the odd customer, which a reply count does not show.
30 build in public ideas, by what you have this week
The examples are made up, written in a short founder voice to show the shape. They are not real user posts. Swap in your own numbers and words before you post anything.
You shipped something
A ship is the easiest post of the week. But "just shipped X" alone gives readers nothing to do. Add the reason, the size, or a job for them.
Ask people to break it
Replies (my rating): High repliesjust shipped dark mode.
find me the screen i forgot.
The request count behind it
Replies (my rating): Medium replies11 people asked for csv export.
shipped it today.
The smallest ship with the biggest effect
Replies (my rating): Medium replies4-line fix.
support emails down by a third this week.
Before and after
Replies (my rating): Medium repliesonboarding: 7 screens → 3.
checking day-2 retention next friday.
Nothing shipped this week
Most founders go silent in quiet weeks. But those weeks have the best questions, because you are stuck on something real.
What you are stuck on, as an A-or-B
Replies (my rating): High repliesstuck on pricing for 3 days.
$19 or $29 for a solo tool?
What you are learning
Replies (my rating): High replieslearning stripe webhooks this week.
what tripped you up first?
The work nobody sees
Replies (my rating): Medium repliesthis week: 0 features, 14 customer calls.
best week in a month.
What you deleted
Replies (my rating): Medium repliesdeleted 2,000 lines today.
product does exactly the same thing.
A number moved
Numbers make building in public worth following. Post the number, the change, and the one thing you think caused it.
A funnel, with a question
Replies (my rating): High replies1,200 visitors 38 signups 3 paid
which step would you fix first?
The weekly revenue number
Replies (my rating): Medium repliesweek 9: $640 → $790 mrr.
the change: annual plan on the pricing page.
The number that went down
Replies (my rating): Medium replieschurn doubled this month.
4 cancels, all said the same thing.
What it costs to run
Replies (my rating): Low repliesmy whole stack costs $61 a month.
hosting is the smallest line.
You made a mistake
People like mistakes more than wins, as long as you share the number and what you changed. Skip the vague "lessons learned" version.
The shutdown or post-mortem
Replies (my rating): Medium repliesshutting down my first product.
5 months. 0 paying users.
what went wrong:
Money you wasted
Replies (my rating): Medium repliesspent $400 on ads.
2 signups. both churned.
The bug users hit
Replies (my rating): Medium repliesbroke login for 40 minutes today.
adding one test so it never happens again.
What you would do differently
Replies (my rating): Low repliesif i started again i would talk to 10 users before writing code.
You made a decision
Readers have opinions about decisions. Post them before you decide when you can, so the replies can still change the answer.
Let them vote on the roadmap
Replies (my rating): High repliesnext feature: team seats or api access?
replies decide.
A pricing choice you are weighing
Replies (my rating): High repliesannual plan: 20% off or 2 months free?
What you said no to
Replies (my rating): Medium repliesturned down a $2k custom build today.
it would have cost me a month.
A price change, announced early
Replies (my rating): Medium repliesraising prices from $19 to $29 next week.
everyone already paying keeps $19.
You talked to a customer
Customer conversations are the source founders use least. Quote them exactly and keep them anonymous. Ask your peers if they hear the same thing.
Ask peers the question your customers ask
Replies (my rating): High repliesfounders,
what is the one question every customer asks you?
The exact quote
Replies (my rating): Medium replies"i would pay double if it did one thing"
a customer, today.
The use case you never planned
Replies (my rating): Medium repliessomeone is using my invoice tool to track chores at home.
The objection you keep hearing
Replies (my rating): Medium repliesmost common reason people do not buy:
"i will just do it manually."
You hit a milestone
On my account, milestones get more likes per view than any other common post type (about 33 per 1,000 impressions). Put the real number and the time it took in the first line. Never round up.
Thank people and invite them to connect
Replies (my rating): High repliesjust hit 1k followers.
if you build in public, say hi and let us connect.
The first payment
Replies (my rating): High repliesfirst paying customer.
$9.
never been this happy about $9.
Day count plus one lesson
Replies (my rating): Medium replies100 days building in public.
0 → 1,400 followers.
biggest lesson:
A question for your audience
These work any week. Word limits and open calls like these got some of my biggest reply counts, because the reader does the talking.
Answer in a word limit
Replies (my rating): High repliesfounders,
describe your week in 1 word.
Pitch me in two words
Replies (my rating): High repliessell me your startup
in two words max.
What are you building
Replies (my rating): High replieswhat are you building this weekend?
drop it below, i will check every one.
If you need more topics than these 30, the 50 post ideas for founders are sorted by proof, customer insight and product decisions. Use this page to pick the shape. Use that one to find more topics.
Build in public examples from my own account
Four real posts, one from each of the strongest groups, with their numbers. All four are short. Each gives readers something to do: congratulate, pitch, cheer or learn from a failure. The first two are shown in full.
First X payout: $828.77 🤯🤯
Took me 81 days to get here from never using X before.
Two lines, one real number, the time it took and where I started. The full story behind it is in 81 days to my first X payout. The post itself did not need any of it.
sell me your startup
in two words max.
The word limit does the work. The reader can answer in three seconds. Every answer is a tiny pitch they want others to see. How word limits did across my account is in how to get more replies on X.
The other two were a number that moved and a mistake. “ClimbX just crossed $1,000 MRR 🤯 50 days since launch. Still feels surreal.” got 135 replies, 241 likes and 7,969 impressions (see it on X). My post shutting down SEOLint, my first X-launched startup, at 0 MRR and 0 revenue ever got 93 replies, 144 likes and 5,383 impressions (see it on X). The shutdown post reached fewer people than the MRR post. But it got the same rate of replies, about 17 per 1,000 views. Founders who had shut something down wanted to say so.
Which build in public posts get the most replies?
Posts that ask the reader for a quick, easy answer: community calls, quick-reply hooks and questions. On my account they were also the most common shapes among my furthest-reaching posts. Posts that give advice were rare among them.
I took my top 500 and bottom 200 original posts by reach (lift against my own median). I labeled them automatically, then compared how often each label showed up in each group. A win ratio of 4.1 means that label was 4.1 times more common among my best posts than among my worst.
- Community call4.1x
- Quick-reply hook3.3x
- Question1.54x
- Opinion0.34x
- Expert advice0.17x
- Playbook0.09x
A study of 500 outlier posts from 181 creators pointed the same way for accounts under 25,000 followers. The replies-per-view breakdown is in how to get more replies on X. Remember your goal, though. This chart measures reach, not trust. A detailed cost breakdown or a long post-mortem can rank low here. It can still be the post that makes a reader trust you enough to buy.
How do you build in public without sounding like an ad?
Talk about what happened, not about the product. On my account, posts that mentioned my own product had a win ratio of 0.40. They were less than half as common among my best posts as among my worst.
Sounds like an ad
ClimbX now has a new scheduling view that makes planning your week easier. Start your 7-day trial today!
Sounds like a founder
spent a week rebuilding one screen.
3 users said the old one made them feel behind.
would you have shipped it or waited?
The second version keeps the product out of the first line. It shares the reason and the real feedback. It ends with a question readers can answer without knowing the product at all.
Let readers find the product on your profile, after they decide they like how you think. Two more results from the same label study: an audience-name opener like “Founders,” had a win ratio of 4.0, and posts of 26 words or more had 0.5. Do not treat that length result as a rule. Across 782 ClimbX accounts under 5,000 followers (last 180 days), posts of 46+ words reached 1.16x their author’s median, slightly better than short ones. Name who it is for, and test length against your own median.
How often should you post build in public updates?
One real update a week is enough to give people a story to follow. Fill the other days with questions, observations and replies. Use the highest-reply shapes rarely, so they stay fresh.
A simple week: 1 to 2 posts where a number moved or you hit a milestone. That is the story people follow you for. 2 to 3 questions, decisions or quiet-week ideas, which get the most replies for the effort. 1 mistake or customer quote, to show you are honest. Community calls at most 1 in 5 posts. They reach furthest, but they ask more than they give.
A winner works once. Repeat the shape, not the post. Repeated winners on my account lost much of their first reach. The numbers behind that are in what is considered viral on Twitter. Use the same idea with a new number, a new audience or a new question instead. To post every week without it eating your day, see how to stay consistent posting on X. ClimbX can also give you a few drafts each day in your own voice, shaped like posts that worked for founders slightly ahead of you. You edit every one, so the week’s real details stay yours.
Before you post a build in public update
- Does it contain one real detail: a number, a quote or a choice?
- Does it give the reader something to do: reply in seconds, save it, or learn a real number?
- Is the product name out of the first line?
- Does every word earn its place, whatever the length?
- Is this the second community call this week? Swap it for a question.
- Is it a repeat of a past winner? Change the number or the audience first.
How these numbers were measured
The label study ranks my original posts on @DanielSmidstrup by lift: a post’s impressions divided by my own median. The win ratio is a label’s share of my top 500 posts divided by its share of my bottom 200. Labels were tagged automatically. I read the reply and impression counts on the four real posts on October 3, 2026. The length comparison is an anonymized total of 52,706 original posts from 782 ClimbX accounts under 5,000 followers with 10+ posts each, last 180 days, my account excluded. The 30 example posts are made up, and their ratings are my own call.
Common questions
What are good build in public ideas?
Post what really happened this week: something you shipped, a number that moved, a mistake, a decision you are weighing, a customer conversation, a milestone, or a question for other founders. For replies, give the reader a job, like voting or answering in one word. For trust or customers, a detailed update or cost breakdown can be the better post.
What do I post when I have nothing to share?
Post what you are stuck on as an A-or-B question, what you are learning, the work nobody saw this week, or a question your audience can answer from experience. Quiet weeks often make the best posts. A real problem gets more replies than a finished feature.
What are examples of building in public?
Sharing a weekly revenue number with what caused it, a funnel with the question of which step to fix, a shutdown with the real numbers, a first payment, or a roadmap vote. On my account, my first X payout post got 717 replies and a two-word pitch prompt got 669.
Which build in public posts get the most replies?
On my account, community calls and quick-reply hooks: shapes that ask for a short answer. I compared my top 500 original posts by reach with my bottom 200. Community calls were 4.1 times more common in the top group, quick-reply hooks 3.3 times and questions 1.54 times. Playbooks and expert advice were rare among my top posts. That says more about reach than about trust or sales.
How often should I post build in public updates?
One real update a week is enough to give people a story to follow. Post questions and observations on the other days. Keep community calls to about one post in five. Do not repost a past winner word for word. On my account, repeated winners got a fraction of their first reach.
Turn this week's work into posts that start conversations.
ClimbX gives you daily drafts in your own voice, built from posts that worked for founders slightly ahead of you. It also finds conversations worth joining. You edit every post, so your week stays the story.
Read next
- How to Build in Public on X: The SaaS Founder Playbook - A complete build-in-public system for SaaS founders: what to share, what to keep private, how to earn feedback, and how to connect posts to customers.
- What to Post on X as a Startup Founder: 50 Practical Ideas - Fifty useful X post ideas for startup and SaaS founders, organized by proof, customer insight, product decisions, lessons, launches, and conversation goals.
- 81 days on X, 6,900 followers, and a first payout of $828.77. - The exact playbook behind 81 days of grinding X: 480 posts, 23,300 replies, 6M impressions, and a first monetized payout of $828.77. Plus why that payout was the least valuable thing it produced.
