Build in Public, and Whether It Brings Customers

Build in public is the advice every founder hears in the first week, and most of them stop doing it by week six. The posts go out. The replies thin to a handful of other founders. The product page gets the same visitors it got before anyone told a story.

The founders who keep going are not more patient than the rest. They decide in advance what they will show, and every post on that list points at something a stranger can act on.

That is the whole argument. Build in public brings customers when the sharing has a destination. It wastes months when the sharing is a diary with no door in it.

Sharing your process costs real hours, and a small brand cannot buy those hours back at a discount. So settle the share list first, send every post to a page, and count the visits. The sections below walk through each step with the numbers filled in.

Build in public starts with a list of what you will not share

Most people who fail at this never wrote the rules down. They posted whatever happened that day. Then they could not explain why a post about a bug fix drew more replies than the post about the new pricing.

Write three lists on one page before you post anything.

1. Things you will show: a shipped change, a decision with its reason, a cost you can name, a mistake and what you changed after it.
2. Things you will keep private: customer names and their data, pay and hiring conflicts, legal matters still open, security holes until they are patched, and any deal still under negotiation.
3. The one page every post can reach: the single address a stranger should land on after reading, written beside the first two lists.

The third list is the one most founders skip. It decides whether the first list ever turns into a sale. A post with no destination is a diary entry, however well it reads.

Work through the table on paper before the first post:

1. Copy the table into a document and add one row for each thing you ship in a normal week.
2. Mark each row share, keep or ask. Ask means the answer depends on a person who has not agreed yet.
3. Pin the document where you write posts. You will check it before every post, for about thirty seconds.

Most of the rows will land in share within a month. The ask rows are where the real judgement sits. Settle them with the person involved before anything goes out.

Share the workflow, not the scoreboard

A useful way to split the practice comes from a 2026 guide at buildinpublic.so. The guide separates performance in public, which means daily revenue screenshots, gratitude posts and every small milestone dressed up as a launch, from workflow in public, which means what you built, how you built it and what it cost. The guide is written by a company that sells workflow tools, so treat its verdict as one view and test it against your own feed.

Testing it costs nothing. Take the next four posts you would have written as milestone posts and rewrite each one as a workflow post: the change, the time it took, the cost, and the link to the page it affects.

Here is the shape a workflow post takes, with numbers filled in. Suppose you rebuilt your signup form last Tuesday and cut it from six fields to three. The post looks like this:

1. Record the old form and the new one in a 40-second screen capture, with the new form on screen for at least 20 seconds.
2. Write the change in one sentence: six fields became three, and the rebuild took about two hours.
3. Write what you expect to move and the day you will check it. For example: "I expect more people to finish the form. I will read the number on Friday."
4. Link the signup page, not the homepage, and tag the link so you can see its visits later.
5. Post it, then reply to the first five comments within an hour. Replies are the one part of the post you can still shape after you hit publish.

The reply window and the link are the two steps most people skip. They are also the two that turn a post into a visit. If you post on LinkedIn, how to get followers on LinkedIn walks through the order of the first month. LinkedIn post ideas gives an opening line for each shape of workflow post. The full set of posts on this topic sits under social media.

Keep the format flat for the first month. Do not add a new platform, a new cadence or a new series until the first four workflow posts have each been read once. Those four posts are the test. Anything else is a second variable on top of the first.

A failure to watch for is the post that shows a result with no reason attached. A chart going up tells the reader that something changed. It does not tell them what to copy. Put one sentence under every number that says what caused it, as far as you know, and what you would try next. That sentence is what a reader takes back to their own product.

Count the link, not the likes

Likes tell you that people liked a post. They do not tell you whether anyone visited the page. Count the visit, and count it per post, because a platform total hides which post did the work.

1. Tag every link you post with a source name for the platform it went out on. The tag is what lets the analytics tell a LinkedIn visit from an X visit.
2. Open your analytics on day one and again on day seven after each post. Write the visits from that post's link next to the post, in a plain list.
3. Set the stop line before you start. Eight posts with fewer than ten visits between them, after four weeks, means the format is wrong, not the platform.

Suppose eight workflow posts went out in four weeks, each with a link. The visits come back as 3, 0, 14, 2, 0, 9, 1 and 0, which is 29 in total. The total clears the stop line. The more useful read is the 14. Find the post that earned it. If it was the screen recording of the signup change and not the milestone post, the next four posts should look like that one.

A total tells you whether to keep going. A single post tells you what to make next. If you want a wider view of which channels send visits in the first place, social media marketing for small business sets out the channels and the order to test them in.

Read what the Buffer case does and does not prove

Buffer is the case founders cite first. Since 2013 the company has published its finances and team salaries, and its open metrics page still lists the December 2013 publication of the full salary list. Business Insider, reporting in 2016, cited a Quartz figure: job applications in the 30 days after Buffer published its salary formula in late 2013 rose about 229%, from 1,263 to 2,886 (Business Insider).

That is a hiring result. It came from a company with a product people already knew and a large audience inside its own industry. It shows that openness can bring applicants. It does not show that a revenue screenshot will bring a buyer.

Read the number for what it measures, and no further.

The other half of the story is that openness has a shelf life. The Bootstrapped Founder's essay Too Many Eyes reports that Buffer later closed its public revenue dashboard, and that Transistor.fm made its revenue private after more than a year of publishing it. The essay gives the reasons its authors saw. Plan for the same outcome anyway. Assume that the day you want the numbers hidden will come, and build the habit of posting workflow now, so that hiding the revenue later costs you nothing.

The buildinpublic.so guide also cites a February 2025 post by Alexander Belogubov that suggested stopping revenue sharing above about $10,000 a month. This post could not open that original post, so treat the figure as one founder's rule and not as a fact.

Before you post a revenue number, write down what a competitor would do with it this week. If the answer is a price cut or a copied feature, post the ratio instead, such as repeat purchase rate or trial-to-paid, and keep the total private. A ratio tells a reader how the business works. A total hands a rival your budget.

Find the point where build in public stops paying

Most posts that fail fail in one of three ways. Name the one you are making before you fix it.

• Stat posts with no link. A revenue figure or a user count with nowhere to click. Fix: attach the page that the number describes, on the last line.
• Gratitude posts. They earn replies from people who already like you and visits from almost nobody else. Fix: replace the thank-you with the thing you learned, and link the page where that lesson applies.
• Volume. One vendor suggests four to seven posts a week. The post could not find a study behind that range, so set your own number and hold it for four weeks before you judge it.

Then run the check at day 14:

1. Count the posts that carried a link.
2. Count the visits those links sent.
3. If three posts in a row sent none, the content gave the reader no reason to click. Rewrite the next one around a result the reader can copy, with the page on the last line.

Build in public does not fit every business, and saying so is part of the practice. Enterprise buyers often expect a confidential roadmap. A hardware company or a regulated one can turn a public date into a liability. If that describes you, do not post the roadmap. Post the thing you ship once it is live, and keep the plans in a private document.

Write one sentence that says what you will share and what you will keep private. Post it once and pin it. Buyers read that line before anything else you publish, and it saves you from explaining your boundaries one reply at a time.

Pay for the attention that build in public cannot reach

Build in public works on people who already know you. A paid placement reaches people who have never heard of you, and the two belong side by side in your plan. They are not competing channels for the same reader. If you want the wider picture of paid options, customer acquisition strategies by budget lays out what each one costs to test.

The Board is a public leaderboard of brands, and The Board runs it. Each brand on it pays to hold a place, and rank follows the amount paid and nothing else. A listing starts at $1, and taking the top spot costs $351 right now. Within a season that price only goes up, until somebody wins the season by holding rank 01 for 30 days.

Read a placement by its clicks, not its impressions. The board models impressions from rank and counts clicks on the outbound link, so the clicks are the number that tells you whether the placement sent anyone to your page.

Run the test the same way you run the posts:

1. Give the placement its own tagged link to the same page your build in public posts point at.
2. Run it for 14 days. Write the clicks from the placement next to the clicks from your posts to that page.
3. Keep whichever channel sent the cheaper visit to that page, and drop the other until the page changes.

The comparison only works if both channels point at one page. Two different pages will tell you about the pages, not the channels. Give the test a fixed budget before it starts, and write down the day you will read it, so the number cannot be argued with afterwards.

Start with the next link you post

Build in public ends where it starts, with a post that points somewhere. Here is the order for this week:

1. Write the share and keep lists, or check the ones you already have against the table.
2. Pick the page your next post will point at, and tag its link.
3. Post one workflow item with a screen capture, the number, and the link on the last line.
4. Reply to the first five comments in the first hour.
5. On day seven, write the visits next to that post, and let that number decide what you make next.

Post the link with the story, then count the visits. The count, not the crowd, tells you what to post next.