Startup pre-launch checklist

Test each of these on a phone, on mobile data, logged out, in a private window. Not on your laptop with your session and your cached assets.

1. A stranger can tell what you do in five seconds. Show the page to someone outside your field and ask them to say what it is. If they hesitate, the headline is the problem, not the design.
2. They can sign up without you. No manual approval, no "email me for access". Every human step you insert is a step you'll have to perform at 2am on launch night.
3. They can pay. Run a real transaction with a real card, then refund it. Test the failure path too — a declined card should say something a human wrote.
4. They can reach you. One email address that reaches a phone you actually look at. Not a form that goes to an inbox you check weekly.
5. You can see what happened. Signups, payments, and where people came from. Three numbers. A full analytics setup can wait; not knowing whether launch day worked cannot.

If any of these fail, you're not ready. If all five pass, you are — regardless of what the rest of your list says.

Your landing page is doing more work on launch day than any feature you might still be polishing. It is the only thing most visitors will ever see.

Check it against what a first-time visitor processes:

• The headline names the outcome, not the product category.
• There is a number somewhere on the page. Anything checkable beats any adjective.
• The price is visible. Not "contact us", not "coming soon".
• The CTA matches the trust you've earned by that point — an email field high up, a purchase lower down.
• There is one line about who it's not for.

Run the free roast below on the live URL and do the first fix it names. It's ranked by leverage, and the top item is almost always the headline — twenty minutes of work for the largest single change available to you before launch.

Enough to not be reckless, not enough to spend a week on. For a solo founder taking online payments:

• Privacy policy and terms. Generated is fine at this stage. Say what data you collect and why. If you have EU or UK users, that's a legal requirement, not a nicety.
• A real business entity if you're taking money. Sole trader is fine in most places to start. Talk to an accountant for an hour before your first invoice, not after your first tax deadline.
• Payment provider terms. Read what your provider prohibits. Founders discover the restricted-business list after a payout freeze, which is the worst possible moment.
• Data location. Know which country your customer data sits in. Someone will eventually ask, and "I don't know" is a bad answer to a buyer.

What you do not need before launch: a trademark, a patent, a shareholders' agreement (if you're solo), an LLC in Delaware, or a lawyer on retainer. Those are real things, and they're for after you have customers.

This is the step that decides launch day, and it's the one solo founders skip because it feels like marketing rather than work.

Two weeks before:

• Write the list. Every community, subreddit, newsletter, group chat and individual who might care. Names, not categories. Aim for thirty.
• Show up in them now. Comment, answer questions, be a person. Posting a link into a community you've never participated in reads exactly like what it is.
• Line up three to five individual conversations with people who'd genuinely want this — for launch day, not before. A personal message on the day outperforms any broadcast.
• Write the posts in advance. Launch day is a bad day for writing. Three versions: short for social, medium for a community post, long for a newsletter or blog.

Then on the day: post everywhere at once, message people individually, and be available to reply for the whole day. Replying is the work. A launch post with fifty comments you engaged with beats one with five hundred views and silence.

Write down what would make launch day a success, before it starts. Specific numbers: "50 signups and 3 paying customers" is a target you can miss honestly. "Good traction" is a target you'll grade on a curve at 11pm.

Also write what you'll do in each case:

• Above target: don't add features. Talk to every single customer this week and find out what they thought they were buying.
• On target: keep the channel that worked and run it again next week. Resist starting a second one.
• Below target: it's a message problem before it's a product problem, roughly nine times out of ten. Go back to the page and the headline before you touch the code.
• Nothing at all: check that you actually reached anyone. A launch nobody saw isn't a failed launch — it's a distribution failure, and it's fixable this week.

Then, whatever happened, relaunch. First launches underperform for reasons that have nothing to do with the product, and the founders who win are the ones who treat launch as a repeatable action rather than a one-time event.