I spent three months building the perfect API. It had clean architecture, 99% test coverage, and a beautiful CI/CD pipeline. I deployed it, leaned back in my chair, and waited for the users to roll in.
They didn't.
I got exactly 0 visitors. Not a single person signed up.
I see this pattern constantly in the #showdev tag. Developers (myself included) spend 90% of their time writing code and 10% of their time thinking about distribution. We think that if we just build a great product, the internet will magically discover it.
But the internet doesn't care about your code. It cares about solutions to problems.
The "If You Build It" Fallacy
As engineers, we are trained to solve problems with code. So when we want to start a side project, our first instinct is to open VS Code and start typing.
We spend weeks building a custom authentication system, a complex database schema, and a beautiful UI. We treat marketing as an afterthought — as if it's "spam" or something that "marketers do."
But building a product without an audience is like building a microservice that nobody calls. It's just sitting there, consuming resources, doing nothing.
The Math of Distribution
Let's look at the math. If you build a product and tell nobody, your reach is 0.
If you build a product and tell 10 people, maybe 1 person signs up.
If you want 100 users, you don't need to refactor your codebase to be 10x faster. You need to reach 1,000 people.
Marketing for developers isn't about spamming Twitter. It's about showing your work and explaining the architecture of the problem you solved.
Write about the bug you fixed. People love a good debugging story.
Share your tech stack. Explain why you chose Next.js over SvelteKit.
Talk about the failure. Post-mortems get 10x more engagement than success stories.
The Shift in Mindset
If you are building a side project this weekend, I challenge you to spend 30 minutes writing the code, and 30 minutes writing a post about why you are building it.
If you're looking for a framework to actually launch your project and get it in front of people, I've packaged my exact launch strategy into a Dev Launch Kit. It's the checklist I wish I had when I was staring at my 0 visitors.
Stop optimizing your package.json. Start optimizing your reach.
What's your biggest struggle with launching side projects? Is it the code or the marketing? Let me know in the comments!
Top comments (9)
That line about 'building a product without an audience is like building a microservice that nobody calls' hits right home!
For me, the hardest part isn't just marketing it's validation before writing code. As developers, building feels comfortable and safe, while putting an idea out there before it's built feels vulnerable.
The 50/50 rule (30 mins coding, 30 mins documenting/sharing) is a great mindset shift to avoid getting stuck in code perfectionism. Solid read!
Your point about vulnerability is spot on. Building feels safe because we control the environment, the compiler, and the logic. Putting an unfinished idea out into the world feels like standing in public without armor—it's terrifying. But as you said, validation before writing code is the only way to avoid building a microservice that nobody calls. The 50/50 rule is exactly the antidote to that fear!
The comparison of a product nobody calls to a microservice with no traffic made this click for me. One thing that worked for me was picking a single community where the problem already gets asked about and answering questions there for a week before posting anything. Did you find one channel gave you real users faster than the others?
Your "answer questions for a week first" strategy is absolute gold. To answer your question: yes, highly niche communities (like specific subreddits or specialized Discord servers) definitely gave real users much faster than broad platforms like Twitter or LinkedIn.
On broad platforms, people scroll past. In niche communities, people are actively searching for a solution to the exact problem you're solving. The conversion rate is night and day. It takes more effort to find those communities, but the payoff is worth it.
hola amigo!!
I disagree with a central point in the article. You say it's not a programming problem; I'd say it's an over-programming problem: too many blocks, too many directives, too many obstacles, patches on top of patches that nobody removes when times change. And if we look at how exposure is designed on the internet, SEO and search engine indexes seem built to hinder the appearance of new repos and articles: today it's easier to drive traffic with a two-minute video than with an article built on testing. On top of that, is there still a community? Active participation seems to die or hide behind closed communities. And the question that keeps nagging me: are we indexing for users or for AIs? Because users no longer browse, they ask their AI, and the AI searches and serves them what they want. The real question is whether it's worth trying to drive user traffic or AI traffic to your repos and articles.
These past few days, living it first-hand, I understood how the rest works. Platforms reward rank, and those who don't have it get blocked for merely breathing (Reddit blocked me on my first article; Google wasn't indexing). Then they wonder why people migrate. About GitHub: its traffic graphs run easily two days behind, verified; and the exposure the platform should give you is left entirely up to you, unless you learn a visibility strategy and apply it. I'm surprised no competitor has emerged to solve this. And about your line: "the internet doesn't care about your code"; I say it would care if it knew it existed. If all repos started from the same baseline, like a game that starts with 1500 points and you go up or down depending on how good you are, how would the system react? A good code publishing system should start from something like that.
And the data backs this up: between 2020 and 2024, posting and viewing content on Facebook and X fell by nearly half, and on TikTok comments dropped 24% while shares grew 45%. The conversation didn't die: it moved to WhatsApp, Telegram, DMs. And that's the problem for newcomers: that private conversation can't be found, can't be measured, and won't bring you a single visit. Before, at least you could go to a subreddit or a Discord and participate. Now the "go to communities" advice becomes harder to execute, because the community is hiding in places where a stranger doesn't enter. Where is a new repo supposed to find its first user, if public space empties out and private space is inaccessible?
Uy, the comment got so long it looks like an article.
You hit the nail on the head with the "indexing for users or AIs" question. We are optimizing for an audience that literally doesn't click anymore. And your point about communities moving to private DMs, WhatsApp, and Discord is painfully accurate—it's the "dark social" effect. It makes the advice "just go find a community" incredibly frustrating for newcomers because the public square is empty, and the private rooms are locked.
Also, that 1500 ELO baseline idea for GitHub is brilliant. The current system is a rich-get-richer dynamic; if you don't have the stars already, the algorithm buries you.
"Building a product without an audience is like building a microservice that nobody calls" - that line is exactly it.
We're building Maqron toward an app that builds apps, and the pull toward "one more feature" instead of "one more post" is real. The feature feels like progress; the post feels like shouting into the void. The reframe that helped us: distribution is also where you find out whether the problem you're solving is the one people actually have. Silence is feedback too - it's just slower.
The 30/30 challenge is a good forcing function. Stealing it.
"silence is feedback too - it's just slower." that is such a brutal but necessary reframe. we always want the instant dopamine hit of a feature shipping, but the silence of nobody using it is the real signal. thanks for sharing that, definitely going to keep it in mind for my own project.
Some comments may only be visible to logged-in visitors. Sign in to view all comments.