Distribution for people who would rather be writing code
You did not build a product in order to become a marketer. The good news is that most of what actually works for a small technical product is closer to engineering than to advertising, and the parts that feel worst are largely optional.
Run the free checkMost of what works for a small technical product is repeatable and system-shaped: making the site machine-readable, publishing one page per week that answers a real question, getting listed in the directories that aggregate products, and being genuinely helpful in a few communities. The parts builders dread most, such as cold outreach and personal-brand posting, are largely optional and are not where the durable results come from.
Why the usual advice does not land
Marketing advice is mostly written by people who enjoy marketing, for people who enjoy marketing. It talks about narrative, positioning, brand, and being everywhere. Read as a builder, it sounds like an invitation to become a different person, and the reasonable response is to close the tab and go back to the editor.
The reframe that helps is this: a small technical product does not grow through persuasion. It grows through being present at the moment somebody has the problem. That is a systems question, and systems are the thing you are already good at.
Almost everything below is a repeatable task with a clear definition of done. None of it requires you to be charming.
The four tasks that matter
Ranked by how much result they produce per hour of discomfort.
- Make the site machine-readable. Real HTML for crawlers, a sitemap, an open robots.txt, unique titles per route, structured data. This is engineering work, it is finite, and it is done once.
- Publish one page per week that answers one real question a stranger would ask. Not about your product. About the problem. This is writing, but it is technical writing, which is the kind you already do in documentation.
- Get listed. Directories, comparison sites, your repository description, the places that aggregate products in your category. An afternoon each, permanent afterwards.
- Be useful in three communities. Answer questions properly, without links, for weeks, until you are a familiar name. This is the only one that requires being a person in public, and it is the one that converts best.
What you can genuinely skip
Worth saying explicitly, because the guilt about not doing these consumes energy that could go into the four above.
You can skip building a personal brand. It helps some people enormously and it is not a requirement, and forcing it when you do not enjoy it produces content that reads exactly as forced.
You can skip daily posting on any platform. The compounding channels do not care about your posting streak.
You can skip cold outreach, unless you are selling to businesses at a price that justifies the time. For a self-serve product it is usually a poor trade.
You can skip most paid advertising at this stage. It teaches you something about your landing page and it stops the moment the money stops.
You can skip video, unless the product genuinely needs to be seen to be understood, in which case one good demonstration is enough and it does not need to become a channel.
Treat it as a system, because it is one
The single biggest advantage a builder has here is the ability to make things repeatable, and distribution rewards that more than it rewards inspiration.
Set a fixed slot. Two hours, one day a week, always the same. In that slot you write the page, or you do the listings, or you answer in the communities. What you do inside the slot matters much less than the slot existing, because the failure mode is not a bad choice of task, it is the month where none of them happened.
Keep a simple list of the questions you could answer, and add to it whenever you see one asked in the wild. Then the weekly page never begins with a blank screen, which is the thing that turns a two-hour slot into a skipped week.
The channel that suits builders best
If you are going to care about one channel disproportionately, make it the AI assistants, because it rewards precisely the things builders are already inclined to do.
What gets a page cited is not charm. It is a direct answer near the top, self-contained paragraphs that survive being lifted out of context, concrete specifics rather than adjectives, and clean machine-readable markup. That is technical writing plus correct implementation, which is a builder's home ground.
It also removes the disadvantage that hurts new products most. An assistant answering a question is not weighing how long your domain has existed. On kurli.me, registered 27 July 2026, ChatGPT sent 56 visitors in the first twenty-five days against 23 from Google Search.
One practical note that costs five minutes: read your robots.txt today and confirm you are not blocking the AI crawlers. Some templates and tools add blanket disallow rules, and a single line can close this channel entirely while you write diligently into a room nobody can enter.
A realistic weekly shape
To make it concrete, here is what two hours a week looks like over a quarter, which is enough to change the trajectory of most stalled products.
- Weeks one and two: the technical pass. Rendering, sitemap, robots.txt, per-page titles, structured data, Search Console. Done once, then not again.
- Weeks three to twelve: one page per week, from the question list. Ten pages by the end of the quarter, which is ten entry points where you previously had one.
- Fifteen minutes of every slot: one directory listing, with a description written fresh for that listing.
- Whatever remains: answer one question properly, in one community, without a link.
Part of a larger guide
This page is one part of Nobody is using it. The other parts:
Questions people ask
- How do I market my product if I hate marketing?
- Focus on the parts that are system-shaped rather than persuasion-shaped: making the site machine-readable, publishing one page per week that answers a real question, getting listed in directories, and being genuinely helpful in a few communities. All four are repeatable tasks with a clear definition of done, which is the kind of work builders are already good at.
- Do I need a personal brand to grow a product?
- No. It helps some people a great deal and it is not a requirement, and forcing it when you do not enjoy it tends to produce content that reads as forced. The compounding channels do not depend on it.
- Should I do cold outreach for my SaaS?
- Only if you are selling to businesses at a price that justifies the hours. For a self-serve product at a low price it is usually a poor trade against the same time spent on pages and listings, which keep working after you stop.
- How much time per week does this take?
- Two hours, in a fixed slot, is enough to change the trajectory of most stalled products over a quarter. What matters more than the choice of task inside the slot is that the slot exists every week, because the failure mode is the month where nothing happened rather than a badly chosen task.
- Why is AI search a good channel for technical founders?
- Because what makes a page citable is technical rather than social: a direct answer near the top, self-contained paragraphs, concrete specifics instead of adjectives, and clean machine-readable markup. It also does not weigh how old your domain is, which removes the disadvantage that hurts new products most.
- What is the one thing to check today?
- Your robots.txt. Some templates and tools add blanket disallow rules that block AI crawlers, and a single line can close that entire channel while you publish diligently into a room nobody can enter. It takes five minutes to read and confirm.
Keep reading
Start with the finite, technical half
The free Get Found Check runs the machine-readable audit for you and tells you exactly which of the technical items are still open on your domain.
Run the free check Have Licheo do it for you