How to Launch a Startup: Launch Early, Then Launch Again

In our view, the most reliable way to launch a startup is to put the product in front of users as soon as it delivers some real value, then keep launching: a landing page, friends and family, strangers, online communities like Hacker News and Product Hunt, and every new feature after that. If nobody notices the first launch, you can launch again.

We think of a launch as a series rather than one moment in time. Each round teaches you something about your message, your channel or your product.

Definition: A startup launch is any deliberate moment when you put your product, or your pitch for it, in front of a new audience in order to get users and learn how they respond. By that definition many successful startups launch dozens of times.

Why we lean toward launching early and often

For many early startups, the bigger launch risk isn't a bad launch. It's a late one. Founders overthink the first launch, spend months polishing it, and then discover that nobody cares. If it took six months to get there, the company may run out of money before it gets a second try.

Three reasons we lean toward going early:

Your theory of the user is still a theory. Conviction is useful, but how people will actually behave stays hypothetical until real people touch the product.
The downside is small. The feared outcomes are that people think it's ugly, a competitor sees it, investors see it too early, or nobody notices. In most cases all of those are survivable. Airbnb launched three times before users started to show up.
Each channel reaches a different audience. Launching in several places tells you whether you're even talking to the right users.

"But you only get one first impression"

This belief is widely held, and it isn't crazy. Founders who worked at large companies absorbed a model in which a launch is a years-long, one-shot event with press in the room. For a big organization with a brand to protect, that model makes sense.

But Try to recall the launch day of Google, Uber or DoorDash. If launches mattered as much as founders fear, you probably would. For a startup, nobody is watching closely enough to remember a rough first attempt, and the one-shot model can be a good way to die anonymous. As a rule of thumb, launch early and launch often.

When to launch your startup

Launch as soon as the product works, which for many teams is sooner than they think. The line we'd draw: a launch is too early mainly if the product adds no value at all, for example if you click the main button and nothing happens. Above that bar, polish, account creation and dashboards are often optional. Brex's early product had no way to create user accounts; the founders set them up by hand.

Two traps can make founders think they've launched when they haven't:

The waitlist that never opens. A founder with a 3,000-person waitlist often counts it as a launch, then balks when asked to let those people in. A waitlist is mostly a form. You get little product feedback from it until people can use the product.
The exception you are not. Founders point to Rippling, which took far longer to launch. But Parker Conrad had already built the same category once at Zenefits, moving fast and scrappy there, so he had earned a big first build. Unless you've done the same thing before at scale, the exception probably doesn't apply to you.

If you don't yet have a working version one, start with how to build an MVP, then come back here.

Write the one-line pitch before any launch

Every launch channel benefits from a sentence that tells a stranger what you do. A few guidelines we find useful:

Lead with what, not why. Company name and what it does first; the origin story can wait for the follow-up question.
No marketing jargon. A listener should have some idea what they would have to build to copy you. A "know-how and synergy platform" falls flat. A line about letting travelers book rooms from locals instead of hotels, close to Airbnb's original application pitch, passes.
One sentence, conversational. If you need a breath in the middle, it's too long.
Use "X for Y" sparingly. It tends to work only if X is a household name, it's clear why Y wants X, and Y is a big market.

Test the sentence on friends before you launch anything. Then rewrite it after each launch based on which words people repeat back to you.

The launch ladder: types of startup launches, from quiet to loud

One way to think about launches is as a ladder. Many early-stage founders climb it in roughly this order, and keep climbing:

Rung	What it looks like	Watch out for
Public placeholder	Domain, company name, one-line description, contact, call to action	Many early startups still lack even this page
Inner circle	Friends and family: watch them use it and ask for feedback	They are rarely your real users; consider moving on fast
Strangers	Go where the pain is and pitch in person	Takes nerve; it is often the fastest signal
Communities	Show HN, Product Hunt, relevant subreddits and forums	Learn each community's rules and tone first
Gated access	Waitlist or request access, with a way to skip the line, often by referral	The longer people wait, the harder conversion gets
Lists and reviewers	Ask writers of "best X" lists to include you	Some are pay-to-play; we would be cautious about paying early on
Pre-orders	Crowdfunding for hardware and physical products	Backers are more skeptical than they were; price carefully
Release launches	Every new release goes out on every channel again	This tends to continue as long as the company is working

The strangers rung is where many of the most useful stories come from. Before DoorDash existed, its founders were building software for small businesses when a Palo Alto macaron shop manager showed them a thick booklet of delivery orders she had no drivers to fill. They interviewed more than 200 business owners, heard the same complaint, and built the first version in a few hours. Lugg's founders parked a rented truck outside IKEA and offered help to shoppers struggling to tie furniture to their cars.

At the top of the ladder, the habit seems to matter most. Stripe launches each new product on Hacker News and answers questions in the thread, then blogs about it and pitches press. Glossier has released new products every six to eight weeks and hit every channel each time.

Press sits last for a reason. Kat Manalac, who headed YC's outreach team and advised batch companies on their launches, makes a point we agree with: a company that has not raised about a million dollars will find US press coverage very hard to land, and even good coverage is not a scalable way to get users or a path to product-market fit.

How to launch a startup on Hacker News and Product Hunt

For many software startups, these two communities are among the highest-value online launches, and each publishes rules worth reading first.

Show HN. Hacker News's Show HN guidelines define a Show HN post as something you made that people can try. Blog posts, sign-up pages, newsletters and lists do not qualify, and landing pages and fundraisers are explicitly off topic. The guidelines ask you to make it easy to try without signups where possible, to be around to answer comments, and not to ask friends for upvotes. Minor version bumps are not enough for a new Show HN; a major overhaul can be.

For the post itself, write like you talk. Introduce yourself, say what you built and briefly why, share something surprising you learned from users, and tell readers what you know well enough to answer questions about. Strong launches aren't only developer tools; posts that feed the community's curiosity do well in any industry.

Product Hunt. Product Hunt's own launch guide has several points to know before you schedule anything:

It suggests launching at 12:01 am Pacific Time if you are planning ahead, and suggests the day you are most prepared is a good day to pick.
You can share your launch link anywhere, but you cannot ask people directly to upvote your product. Ask for visits and comments instead.
It encourages makers to hunt their own products and sees no discernible advantage in a third-party hunter.
You can relaunch whenever you have a significant new product iteration, which fits the launch-again model.

Which converts better? YC has compared how launches on TechCrunch, Product Hunt and Hacker News convert into users and customers, and the results have been evening out. Our read: pick the channels where your users actually are, not the one with the most prestige.

Our own launch. 1752vc launched its Pitch Deck Analyzer on Product Hunt on a Saturday and finished as #1 Product of the Day with 352 votes and 84 comments, 139 votes ahead of the runner-up, as our September 1, 2026 write-up, "You Can't Buy #1," reports. The groundwork took roughly 40 days: a waitlist emailed personally in waves segmented by timezone, a listing refined like a pitch deck, and a habit of not asking anyone for an upvote. The biggest lesson we took from it: comments, not votes, won the day, which matches Product Hunt's own push for conversation.

Worked example: a 12-week launch calendar with funnel math

A two-founder B2B startup has a working MVP for invoice reconciliation. Instead of one launch day in week 12, it plans a launch every one to two weeks. The figures below are illustrative assumptions, not benchmarks, and real results vary widely.

Week	Launch	Reach	Signups
1	Public placeholder: landing page with one-line pitch	Links in every email	Collect emails
2	Friends and family	20 people	20
3 to 4	Strangers: 30 direct conversations with finance leads	30	30
5	Show HN	3,000 visitors at 6%	180
7	Two niche finance communities	1,200 visitors at 10%	120
9	Product Hunt	2,500 visitors at 7%	175
11	Feature launch to the email list	525 on the list	Reactivation

The three public launches produce 475 signups. Now apply a filter we find useful in early B2B sales: the job is less about convincing everyone and more about sorting 100 signups down to the five or six with a hair-on-fire problem. At 5 percent, the 475 signups yield about 24 qualified prospects. Add roughly 4 from the stranger conversations and the founders have about 28 people worth a real sales conversation.

If 30 percent of those convert to paying at $200 a month, that is about 8 customers and $1,600 in monthly recurring revenue by week 12. By then the list holds everyone from every launch, 525 people, so the week 11 feature launch reaches about 210 readers at a 40 percent open rate.

Compare the single big launch. One Show HN in week 12 with the same numbers gives 180 signups, about 9 qualified prospects, and no chance to fix the pitch, the page or the product between attempts. The calendar didn't just add traffic. It gave the founders seven rounds of feedback on their message and their product.

Notice where every signup ends up: on a list you own. Platforms come and go, and their reach fades over time, which is our argument that channels decay rather than die. The email list is the one audience you keep between launches.

Launch-day checklist, post template and mistakes

Before each launch:

[ ] The product delivers its core value without the founders explaining it live.
[ ] Your one-line pitch leads with what the product does, with no jargon.
[ ] The landing page has a clear call to action and a working signup.
[ ] You know each community's rules (Show HN excludes sign-up pages; Product Hunt bans asking for upvotes).
[ ] Analytics track visits, signups and the first key action.
[ ] A founder is free to answer comments for most of the day.
[ ] You have a plan to talk to qualified signups within about 48 hours.
[ ] Signups land on your own email list for the next launch.

One template for a community launch post:

Title: Show HN: [Name], [what it does, in plain words]

Hi HN, I'm [name]. I built [product], which [does what] for [whom].

Why: [one or two sentences on how you ran into the problem].

What surprised us: [one real insight from users or from building it].

Try it: [link, ideally with no signup required].

It's early and rough in places, especially [known gap]. I'd love
feedback on [specific question]. Happy to answer questions about
[the areas you know deeply].

Common launch mistakes:

Waiting for a perfect launch day. First-time founders often take too long.
Asking for upvotes. Both Hacker News and Product Hunt rules prohibit it.
Sitting on a waitlist. Conversion drops the longer people wait.
Paying for launch placement. We'd be cautious about paying bloggers or agencies early on; for our own launch we turned down paid agency and vote-selling offers.
Launching once and stopping. Each feature can be another launch.

When you launch your startup and nobody shows up

It happens to many founders. Treat it as a diagnosis, not a verdict. Find where people drop off. If prospects aren't replying, change the message. If the message is fine, maybe the target is wrong, for example selling to thousand-person enterprises when smaller companies feel the pain. Change one variable at a time, week by week.

If nothing works after a real stretch of trying, ask which of your assumptions was wrong. That doesn't automatically mean the idea is bad. It means you need a new hypothesis before the next launch.

Many founders find it helps to set the goal of each launch as learning rather than a revenue number. That makes it hard to fail and much easier to ship. The team behind COLOURlovers, later Creative Market, once staged a big relaunch with reporters in the room, took the site down under a flood of search engine indexing, and found that nobody much cared either way. Take the pressure off any single launch.

Turning a stream of launches into paying customers is its own skill, and the one 1752vc's GTM Accelerator is built around: a 12-week, hands-on, remote and self-paced program for founders who have validated their product and have early traction, teaching them to sell, recruit, fundraise and build traction, with rolling admissions. The follow-up conversations after each launch are where founder-led sales starts, and many strong early users come from doing things that don't scale.

Where views differ on launches

One channel or every channel? Some founders swear by being everywhere and planning a launch for every community they belong to. Lenny Rachitsky's 2020 study of how major consumer apps found their first 1,000 users points the other way: most found their early users through a single strategy, a few used a handful, and none succeeded through more than three. Our take: use many small launches to find your one or two channels, then double down on them.

Launch day as an event. Product Hunt's guide and our own launch both treat launch day as something to prepare for, with timing, assets and weeks of community work. That can look like it contradicts launching often. We don't think it has to: prepare well for each individual launch, but don't let preparation delay getting the product into users' hands, or treat any one launch as your only shot.

Waitlists. Waitlist launches have worked, as with Robinhood and Superhuman, and Dropbox's demo video drove 75,000 waitlist signups in 24 hours by TechCrunch's 2011 account. We still don't count a waitlist as a launch until people can use the product. Use it to capture demand, then open it fast.

Once launches bring in users who stay, the question shifts to whether you have product-market fit.

The bottom line

Ship when the product delivers real value, not when it looks finished. Then launch again on the next channel, with a sharper one-liner and a list that grows each time.

One launch is a bet. Ten launches are a feedback loop.

Key takeaways

In our view, it usually pays to launch as soon as the product delivers some value, then keep launching on new channels and with new features.
A launch is usually too early only when the product does not work at all; missing polish, accounts or dashboards is rarely a reason to wait.
It helps to write a one-line pitch that says what you do, for whom, with no jargon, before any launch.
On Show HN and Product Hunt, follow each community's published rules (both prohibit asking for upvotes) and consider spending launch day in the comments.
One way to treat each launch is as a filter and a diagnosis: find the few users with an urgent problem, fix one variable at a time, and launch again.

Frequently asked questions

When should you launch your startup?

In our view, as soon as the product delivers its core value, even if it is rough. A launch is usually too early only if the product does not work at all, for example if the main button does nothing. Missing polish or features is rarely a reason to wait, because each week unlaunched is a week without user feedback.

How do you launch on Product Hunt?

Product Hunt's own guide suggests launching at 12:01 am Pacific Time on the day you are most prepared and hunting the product yourself. Prepare your listing, tagline and gallery in advance. Share the link widely but ask for visits and comments, not upvotes, which Product Hunt prohibits. Many makers spend the day answering comments.

How do you post a Show HN on Hacker News?

Submit a post whose title starts with "Show HN" linking to something people can actually try, not a sign-up page, blog post or landing page. The guidelines ask you to make it easy to use without signing up; it also helps to write plainly about what you built and why, and stay in the thread to answer questions. Hacker News rules forbid asking friends to upvote.

What should you do if nobody uses your product after launch?

One approach is to treat it as a diagnosis. Find where people drop off, then change one variable at a time: the message, the channel or the type of customer you target. Ask which of your assumptions was wrong and form a new hypothesis. Then launch again; Airbnb launched three times before users showed up.

Should a startup launch with a waitlist?

A waitlist can capture demand, and companies such as Robinhood and Superhuman used them well, but a waitlist is not a launch because nobody is using the product yet. Conversion also gets harder the longer people wait. We would lean toward opening access quickly, and letting early users skip the line through referrals so the list grows while you onboard.

Sources

Disclaimer: This guide is for general education only and is not legal, tax or investment advice. Laws, market data and program terms change, so it may not reflect the latest developments or fit your situation. Treat it as a starting point, not a source of truth, and talk to a qualified lawyer, accountant or financial adviser before you make decisions.