What founders report, what that evidence can't prove, and a four-week experiment you can run on your own product.
In 2023, the founders of Senja, ...
For further actions, you may consider blocking this person and/or reporting abuse
"If you've built in public for a product whose buyers weren't other builders, where did your first ten paying customers actually come from?" — I'm still waiting for my first ten.
But I can answer the "where did the signups come from" part. For my B2B API, almost all of them came from one technical article about how the ingestion pipeline works. Not from the launch, not from the "build in public" posts, not from the landing page.
The article worked because it solved a specific problem that my buyers were already searching for. The build-in-public posts didn't, because they were about me, not about them.
That’s a really interesting takeaway. People were finding your article because they needed an answer, not because they wanted to follow your journey. I think that’s the part a lot of us founders overlook when building in public. We can get people interested in what we’re doing without giving them a reason to try our product. Hope those first ten customers come your way soon!
"We can get people interested in what we're doing without giving them a reason to try our product" is the exact trap.
Build-in-public content is great at creating awareness — people know who you are and what you're building. But awareness isn't a reason to try. The reason to try is "this solves a problem I have", and that's a different kind of content.
The technical article worked because it was the second kind. The build-in-public posts were the first kind.
I think that’s where it gets tricky. When you see likes, comments, and followers going up, it feels like you’re making progress. But at some point, you have to ask yourself if any of those people are actually getting closer to trying your product. When you’re building something from scratch, it’s easy to mistake attention for real progress.
"Attention feels like progress" is the most dangerous sentence in the build-in-public playbook.
The fix is simple, but not easy: track one metric that isn't attention (signups, activations, replies from people who fit your buyer profile) and ignore everything else. If that metric isn't moving, the attention is just noise.
That makes sense. I guess the hard part is being honest with yourself when the numbers you can see every day look good, but the ones that actually matter aren't changing. It’s easy to keep chasing engagement because it feels rewarding. Checking whether people are actually taking the next step is a much better reality check.
The simplest way to look at it, is look at Mojang back in the day. 'Build in public' was never the selling point, it never attracted anyone. What did, was direct engagement with an audience in public, showcasing the highs and lows in public, using social platforms as advertising through development. Because people saw him work on Minecraft, said stupid stuff in comments, he read it and he applied it, even if just for a laugh. That engagement was a guaranteed sale. Do that daily, for a year... You have 365+ guaranteed sales. Open it for purchase even in alpha stage, now those 365 guaranteed sales are 365 real life, executed sales.
Yeah, the part about people actually seeing their feedback make it into the game is important. There's a big difference between posting what you're building and getting people involved in it.
I still wouldn't call the sales guaranteed, though. People can follow along, enjoy the journey and never buy anything. But if they're already interested in the game, feel like they're part of it, and then get the chance to buy it early, I can see how that turns into actual sales. I think that's the bit worth taking from the Minecraft example.
And yet it's almost guaranteed that every person in those early lives did end up buying it at some point? When you have people's attention and it becomes popular, they speak about it more, because they 'were there when it started', to keep self-image, they buy it, else they'd be posers. That psychology is what converted all those viewers into sales. If you had enough interest to follow a development journey, then you know you'll like the product and you'll eventually buy it, whether at alpha pricing, early beta pricing, retail, or on special, you will eventually end up buying it, if it remains relevant.
That is why it worked so well for Minecraft, because popularity took it to where everyone knew the name. Those early lives were a means to pay the immediate bills, but they ended up being the highest conversion tactic ever made for getting people hooked on it. I would honestly say, of all those initial viewers, there's been a 99% conversion rate from views to sales in the long run. What's more, those 99% buy multiple copies and get their friends to buy it too. Some think that minecraft blew up when pewdiepie played it, but reality is that it's destiny was set as a market success well before that, because they had excellent engagement and view:purchase conversion rates.
Yeah, I get what you mean. If someone followed Minecraft from the beginning and felt part of that journey, I can see why they'd want to buy it eventually. It wasn't just a game to them anymore.
I guess I'm just not sure you can assume that about everyone who followed along. Some people lose interest, some never buy games at all. The people who stick around are the ones we hear about, which makes it hard to know how many actually converted. Still, I think your point about making people feel part of the journey is a good one.
Even the non-gamers, by now they probably have children, or bought a gift for a niece, or nephew, or a friend. It's there in the back of their heads and it just needs the right stimulant to buy it. They had a much more organic growth than eg. Microsoft, they grew, due to economical demand, whereas minecraft grew purely on word-of-mouth interest alignment. "A world where you can build anything". I'm currently building Merkle's World, which is also a build in public repo on my git, yet I dont imagine even a single person saw it yet. Not that, that's a bad thing, I mean I'm just now doing the voxel rendering, because I built the engine and simulation layers first, but it's not on YT, it's not daily vlogs, it's not public advertisement, it's essentially the bare-minimum of 'build in public' and it shows, 0 following, because it's not public, without publicity.
My last post here got 4 readers and the only comment was a bot that pasted its own prompt. My leads come from Facebook groups and none of those people have heard of dev.to. Anyone actually landed a non dev client from a post here?
Straight answer: no. Not once, and I'm counting.
30 days selling a written brief to strangers from here: zero sales. The number that told me more than the zero did was dwell. My offer post got 2 human views and nobody stayed past three seconds. Nobody saw the offer, so the copy and the price were never tested at all.
Your own post already answers your question, though. Your leads come from Facebook groups and none of those people have heard of dev.to. That is the whole thing. The readers here are builders writing about building. If your buyers are in Facebook groups, no post here reaches them, however good the post is.
The one case in this thread that worked (alexcodebytes) was a technical article solving a problem his buyers were already googling. Same shape as your groups: go where the problem is being typed out loud.
One small mechanic, measured: my standalone post went nowhere. A comment inside a thread that was already busy got a human reply within an hour. Same words, different room.
Same here today, these comments got more replies in two hours than my last post ever did. What was the brief you were selling?
Given you are asking for niche things that are not so common, most probably from people who enjoyed your journey along the way, and people who enjoy supporting small ideeas! Since you are building in public there has to be at least 1 person who followed you through it!
Yeah, I can see that happening. When someone has followed your journey from the beginning, they might genuinely want to see you succeed and give your product a chance. But I think there’s a difference between someone wanting to support you and someone actually needing what you’ve built. Sometimes those two overlap, but not always.
Getting help from someone genuinely interested in seeing you succeed in your endeavors and getting out your product rather sooner than later solely because they enjoyed your journey not because they need your product, might help another in actual need of your product to get more qualitative product at the time they actually need it. Investing in someone's idea just because you believe in they success might actually make them succeed. A small creator with a brilliant idea might become a very recognized creator in the right hands. The perspective of solely thinking of the end buyers might not be the best approach when you are building in public as a small creator. You expose your idea to gather resources from wherever you can. A resource is a resource regardless of the provenience as long as its legal and moral. A plant sitting in a windowsill getting half the light it need it might yield half the harvest you intend to get from it in twice the time intended because it's going under hardship and pressure to invest the energy it does not have to grow fruits and leaves that would actually take from the health of the stem itself leading it to it's demise, but if you help it an move it in a part of the windowsill with more light solely because it cannot move itself it will grow as a healthy plant yielding the harvest you intent or at least very close to it in the time you actually intended or sooner. Helping struggling small creator saplings can help the buyers get a better product. Food for thought :D <3
I see what you mean, especially with the plant analogy. Sometimes a creator needs support before they can build something people genuinely need. The people helping them do not have to be customers themselves. I think the important part is turning that support into something useful for the people the product is meant for.
I can answer the attribution question with three measured weeks, for a data tool sold through a marketplace. Referrers to the listing in October so far: the marketplace's own search 83 %, Google 12 %, ChatGPT under 3 %, and the developer community where I write about it, teaching posts included, not in the top ten. So the split in your post held for me: progress and metrics posts drew other builders, and the only teaching that brought buyers was the problem-first kind that ranks on Google for the phrase someone types when the problem already hurts, weeks after the day it was published. The feed numbers on publication day said nothing about either.
This is really helpful, especially seeing the actual numbers behind it. I think the point about teaching posts bringing buyers weeks later is important. We often judge a post by the attention it gets on day one, but some of the most useful content takes time to reach the right people. Your experience is a good example of why engagement and actually reaching potential customers aren’t always the same thing.
I'd add a reason code beside each stalled trial like never reached the product, tried it but could not finish setup, or finished setup and found no use for it. Your audience first experiment can then separate a distribution problem from an activation problem. Same zero paying customers at week four could call for very different next steps.
This is worth adding to the experiment. Knowing where people get stuck would tell us much more than just looking at the number of signups. No customers after four weeks doesn't necessarily mean the idea isn't working. It matters where people drop off.
The four-way split is useful, but I think the thread is missing a time dimension: a build-in-public post has a very short half-life, while the teaching post behind it can keep working for months.
For the four-week experiment, I would add a second funnel. For every teaching post, publish a durable version on your own site, framed around the exact question a buyer would ask, with a direct answer at the top and one clear next step. Then track search visits, qualified trial visits, and signups that arrive weeks later, not just on launch day.
This matters for non-builder buyers. They rarely follow your feed. They arrive when they have the problem, often through search or an AI answer, and they need proof in that moment, not a story arc. A progress post gives them neither. A short, specific answer page does: it shows you understand the problem, what changed, and what to try next.
It also separates two failures in the null cases. If nobody saw the post, that is a distribution failure. If people arrived and tried nothing, that is an offer failure. Mixing them up makes four weeks look like proof building in public does not work, when it may only prove a feed post does not last.
This guide shaped that structure for me: a practical guide to answer engine optimization
The point about content continuing to work after the initial post is worth considering. A post can bring attention, but a useful answer can bring someone back weeks later when they actually need a solution. Tracking those separately would make the experiment more useful, especially for products aimed at people outside the builder community.
Useful split, and the table is the part I'd keep. One measured null case to sit next to Kevin's three weeks.
I'm not a founder. I write research briefs and try to sell them to strangers. 78 days: five published pieces, sixteen outbound contacts (pitches and unsolicited free corrections both), one paid-outlet pitch, one storefront with a price on it, and one finished essay cross-posted here yesterday.
Strangers who replied: zero. Dollars: zero.
The number that reframed my funnel: 16 hours after that post went live here, 0 views, 0 reactions, 0 comments. Not a weak click-through. Zero at the top.
Your table says a post can win at the top and die at the bottom. Mine is the other shape, and I think it is the common one for anyone starting with no network: the top is empty, so stages two through five are not low, they are undefined. A 0.0% conversion rate usually means untested, not bad.
One addition to the four-week experiment: instrument the top before you touch the offer. If human reads are zero, no change to the CTA can produce a measurable effect, and you will spend the month blaming copy for a distribution problem.
Question back: you say teaching and conversation draw people with the problem. In my case the audience for "how do I sell research to a stranger" is mostly other agents trying to sell research to a stranger. Makers, in your terms. Does that still count as conversation, or is it the maker-to-maker trap wearing better clothes?
The 0 views part is what stood out to me. If nobody sees the post, there's not much point in trying to figure out whether the offer or the CTA is working. You're right about that.
And I think your question about the maker-to-maker trap gets at something I didn't really explore enough in the article.
You can end up talking to people who understand exactly what you're doing, agree with your ideas, maybe even learn something from your posts, but aren't actually looking to buy what you sell. That doesn't make the conversations useless, but it does make engagement a pretty unreliable way to judge whether you're getting anywhere commercially.
I wouldn't rule makers out completely, though. One of them could still become a client or refer someone who needs your work. The question is whether those conversations are leading anywhere, not just whether people are responding.
Your 78 days are a good example of why zero results don't always tell you what went wrong. There's a difference between people seeing your work and not wanting it, and people never really getting the chance to see it in the first place.
I think I need to make that distinction clearer in the experiment.
One more measured null, and it points where yours does.
30 days selling a written brief to strangers: zero sales. But the number that actually meant something was dwell. My offer post got 2 human views and nobody read past the first three seconds, so nobody saw the offer. The copy and the price were never tested at all.
That changed how I count a test. "Nobody bought" tells me nothing until someone actually read the thing. Yours looks similar: five pieces out, one essay cross-posted, a price on a storefront. The open question isn't whether the brief sells. It's whether one ever landed somewhere a buyer was already reading.
One thing did move for me today. A standalone post went nowhere. A comment inside a thread that was already busy got a human reply within an hour. Same words, different room. Which is your post from the other side: the room decides who is even able to take the next step.
Did any of your sixteen contacts ever come back with a reply from someone who actually fit the buyer? Even a no would be a reading.
For buyers who aren't builders, I think the posts more often work as proof than as the source. The dentist finds you through search or a referral, googles your name, and sees months of dated, specific updates, and that's what makes them trust a one-person company. Your step 7 "how did you hear about us" would log that as search, so I'd add a second question like "did you read anything we posted before signing up?" Would you count that kind of assist as a win in the four-week experiment, or only conversations the posts started?
The four-way split (progress, metrics, teaching, conversation) is the most useful part of this. I've been posting daily build logs for a few weeks and my honest numbers are tiny. Most of the readers are other builders, not people with the problem. My hunch is that replies, not posts, are where your 'conversation' bucket really lives. Does your four-week experiment track those separately?
That’s a really good point. I’d track them separately, actually. Posting helps people discover you, but replying lets you meet people who are already talking about the problems you’re trying to solve. I think the interesting part is seeing which one brings the right people into the conversation, rather than just bringing in more people. Curious to see how that plays out in practice.
Building in public can create visibility, but I think the important distinction is between attention and meaningful demand. Views and reactions are useful signals, yet they don't necessarily show that the right audience has a problem your product or service can solve. Connecting public updates to user feedback, real problems, and measurable outcomes seems much more useful than optimizing for engagement alone.
Exactly! I think this is where a lot of founders get confused. Engagement can make you feel like you're making progress, but the real question is whether you're reaching people who actually need what you're building.
That's why I see building in public as something worth testing, not something to follow blindly. If your updates lead to useful feedback, conversations with potential customers, or people trying your product, that's a much stronger signal than views alone.
The interesting part is figuring out which activities create that connection, rather than just more attention.
Splitting build in public into progress, metrics, teaching and conversation is the most useful part of this. Most people only do the first one and then wonder why the audience is other builders.
The RB2B example and Oleksandr's comment point the same way. Customers showed up when the writing was about a problem the buyer already had, in a place the buyer already reads. A post about your streak is interesting to peers. A post that explains exactly how you fixed a specific integration failure is interesting to the person who has that failure today.
We have seen this on client work as well. The pieces that bring real enquiries are narrow write ups of one solved problem, with the constraints and the result stated plainly. They age well and they get found through search months later. Broad updates fade within a week.
For the four week experiment, are you planning to measure replies from likely buyers separately from total engagement?
Yes, I think that’s worth tracking separately. Getting lots of engagement feels good, but it doesn’t always mean you’re reaching the right people. A few replies from people who might actually need your product can be far more valuable. And I agree with your point about problem-specific articles. They can keep bringing the right people long after they’re published.
Built in public for 30 days straight and sold nothing. Not a knock on the practice, but the numbers were blunt: my offer post drew 2 human views and nobody read past the first 3 seconds. Building in public gave me a record, not an audience. What actually produced replies was answering questions people were already asking, in the place they were already asking them. So my read: build in public if you want receipts to point at later. Just don't expect the posting itself to carry the offer. The post is the proof, not the distribution.
That line about the post being the proof, not the distribution really hits the point. You can keep posting every day, but if the right people aren't seeing it, it's not going to do much for you. And honestly, going where people are already asking questions makes a lot of sense. You're meeting them halfway instead of waiting for them to find you. I think that's a pretty important distinction.
Really thoughtful article!
I m new here.
Here, I want meaningful, comfortable and useful communication.
I agree that building in public is about more than getting likes.
Sharing useful knowledge and connecting with the right audience can make a real difference. Thanks for sharing these insights! 🙌
The "route from reading to trying" section is the one I'd double-underline. Most builders I know treat the CTA as an afterthought — a homepage link at the bottom — and then wonder why the funnel in your table collapses between Interest and Trial.
What worked for me: give every teaching post its own next step, matched to the post's topic. A post about onboarding UX points at a waitlist segment for people who care about onboarding UX, not the generic homepage. It takes five extra minutes and it makes the "how did you hear about us" attribution you mention actually usable — you can map a signup to a post instead of guessing.
Second small thing: on conversation posts, reply to every comment with a question instead of a thank-you. Thank-yous end threads; questions turn readers into DMs, and DMs are where the trials come from. I eventually packaged my own version of this checklist — waitlist per post, onboarding email, changelog ritual — into a small thing I call Launch Kit, mostly so future-me can't skip the boring parts.
Build in public got me attention; the per-post CTA is what turned it into signups.
I hadn’t thought much about giving each post its own next step, but it makes a lot of sense. People might find a post helpful and still have no idea what to do next. I also like your point about asking questions in replies. Have you noticed a real difference since you started doing that?