How to launch a product people can find

Sunday night. You hit deploy, and it works. Signups, checkout, the little success animation you rewrote three times. You send the link to a friend and go to bed feeling like a founder.

Monday: eleven visitors. Six are you. Two are your friend. The rest look like bots.

Nothing is broken. Your product lives on a brand-new domain, and a brand-new domain is an empty side street at 3 a.m. You hung the best sign in the world over the door. Nobody walks by, because nobody knows the street exists yet. Getting a street known takes months.

A launch is how you survive those months.

The short answer: don't wait for Google. Take your product to streets where people already walk. Make the offer readable (the sign). Show it to a crowd on launch day (the bell). Leave listings behind that drip traffic for years (the cup). Then do it again with the next product, and keep whichever one pulls.

That's the whole plan. The rest of this page is the plan in order.

Why nobody finds you yet

You did nothing wrong. Search engines don't trust streets they've never seen. A crawler stops by, reads a few pages, and leaves. Rankings come later — after other sites link to you, and after people start finding you on their own. That machine is slow by design.

People in SEO have a name for the slow start: the "Google sandbox". The idea is that a new site sits near-invisible for its first months no matter what you do. Google has never confirmed a sandbox exists. What anyone who has launched will confirm: a fresh domain waits.

So don't wait with it.

The cup, the bell and the sign

Picture a street vendor. Three tools, three jobs.

The sign is your offer. One line that says what this is and who it's for. If the sign is unreadable, everything after it is wasted.

The bell is a launch platform. A bell works because a crowd is already standing in that square. Ring it on the right day and hundreds of people look at once.

The cup is a directory listing. Small coins, dropped in slowly, for years. Not a flood. A drip you set up once.

Most people who "do marketing" after shipping grab only the bell. The order that works is: fix the sign, then ring the bell, then set out the cups — and talk to people face to face in between.

Fix the sign first

Before Product Hunt, before Reddit, write one sentence:

What it is, for whom, so they can what.

No adjectives. No "revolutionary". "Invoicing for freelance translators" beats "the future of payments" every time a stranger reads it.

The test: show the sentence to someone outside your head. If they ask "so what does it do?", rewrite it. If they say "oh, my sister needs that" — done.

A few made-up before-and-afters, to show the shape:

"AI-powered productivity platform" → "Turns your meeting recordings into a to-do list"
"The future of fitness" → "10-minute workouts for people who sit all day"
"Smart analytics for creators" → "Shows which of your videos brought paying subscribers"

Then make the page match. A visitor gives you about five seconds, and they're deciding three things: what is this, is it for me, what does it cost. A 30-second demo video or a short GIF does more than a page of copy. When a launch page fails, it's often this — an unclear sign, not a bad product.

Ring the bell: launch platforms

A launch platform is a square where a crowd already gathers. You bring the product to them, on their schedule.

Product Hunt. The classic first bell for software. One day, upvotes and comments, a rank on the front page. It has its own rhythm — tagline, screenshots, maker story, and showing up all day to answer every comment. We wrote a full walkthrough: How to launch on Product Hunt.
Hacker News. A "Show HN" post for anything technical. Smaller odds, sharper feedback. The comments will tell you what your sign failed to say.
Communities in your niche. A subreddit, a Slack, a Discord, a forum where your exact users already talk. Often better than the big squares, because everyone there is your person.
App stores, if you built mobile. The store is its own square with its own rules — screenshots, keywords, reviews. Worth an afternoon, not an afterthought.

One honest note on press. People used to "submit a startup" to big tech blogs and hope. For an unknown maker, that's a lottery ticket. Send the short pitch if you like — it costs an email — but don't build your month around it. The bells above are ones you can actually ring yourself.

Fill the cup: directories

Directories are the cup. A listing takes ten minutes and then drips: a bit of traffic, a place where people browse for tools like yours, and a link from another site to yours — a backlink, which search engines count roughly like a vote.

Not every directory earns the ten minutes. Some are dead. Some charge money for a listing nobody sees. Some are link farms — sites that exist only to emit links — and Google's spam policies treat bought and mass-produced links as spam, so a bad listing is worse than none.

We went through which directories are worth it and which to skip: Startup directories: which are actually worth submitting to.

Talk to people where they already are

Between launches, be a person. Find the three places your users gather — a subreddit, a Slack, an old-school forum — and show up before you need anything. Answer questions. Share what you learned building the thing.

Then, when you share your product, share it where it's relevant, once, plainly. One genuinely useful answer in a thread can outpull a directory listing for months. Spam ten threads and you'll get banned and remembered. Slow is fast here.

"Zero cost" is mostly true

You'll see people search for zero-cost startups, and the launch part really is close to free. A domain costs about ten dollars. Hosting has free tiers. Launch platforms and directories don't charge. The real budget is hours: writing the offer, preparing the page, answering comments, submitting to directories, being useful in communities. Budgeting for a launch mostly means budgeting time — a focused week, not a focused wallet.

Built it without code? Doesn't change the plan

Plenty of products now get built in a weekend with no-code tools or AI assistance, and that's fine. The market doesn't care how it was built. The market cares whether the sign is clear and whether the product solves something.

What cheap building does change is strategy: if testing an idea costs a weekend, the smart move is to test several. That's the last piece.

The launch is the validation

People ask how to validate a business model. A launch is the cheapest honest test there is.

Upvotes and comments mean the offer is interesting.
Signups mean it's interesting enough to try.
Credit cards mean it's validated.

Silence is data too. If a hundred strangers see your sign and nobody bites, the product or the sign is wrong — and you found out for the cost of a weekend, not a year. Rewrite the offer and launch again, or move to the next idea.

This is the part most people get backwards: they pour months into one product and then mass-produce web pages to catch searches for it. Don't. Ship many products, keep what works. Mass-producing pages for search is what Google's spam policies (2024) call scaled content abuse — flooding the web with low-value pages to game rankings — and it can sink the whole site. We made that argument in full here: Build products, not pages.

Building on a domain that isn't fresh? Check its past

Some people skip the slow start by building on a domain that already has history. That can work, and it can burn you.

A good expired domain is a corner that signposts have pointed to for years — links from newspapers, universities, encyclopedias. People, and Google, already walk there. A bad one is a corner where stolen goods were sold: there's foot traffic, but the police are watching — spam, casino, pharma somewhere in its past.

Google's spam policies also name expired domain abuse: buying an old name to host unrelated, low-value content and ride its old rankings. In plain words — if the history doesn't match what you're selling, you inherit the suspicion along with the links.

Before you build on a second-hand domain:

Look it up in the Wayback Machine. What lived at that address for the last ten years?
Run a site: search on Google. What does Google admit it still has indexed?
Check Google Safe Browsing for warnings.
One thing nobody outside Google can see: whether a domain has a manual penalty. Only Search Console shows that, and you need to verify the domain first. Anyone selling you a "guaranteed clean" domain is guessing.

And if the name might collide with someone's trademark, talk to a lawyer before you print the sign. An hour of legal advice is cheaper than a rebrand after launch.

What to do now

Write the one-line offer. What, for whom, so they can what. Test it on one human outside your head.
Make the page pass the five-second test. What is it, who is it for, what it costs, one demo GIF.
Verify Google Search Console (free) today, so when search traffic does start, you can see which queries bring it.
Pick one bell and one community. Read their rules, prepare, launch, and answer every comment that day. Product Hunt is the standard first bell for software — here's the full playbook.
Spend an hour on directories that earn it — our shortlist is here — and if you're choosing a second-hand domain, check its past first: the Wayback Machine, a site: search, Safe Browsing, or our DD Backlink Strainer Chrome extension, which vets domain lists on ExpiredDomains and major auctions in a side panel (free plan: 250 checks a day).

The street gets known one visitor at a time. Ring the bell while it does.