MVP Launch Checklist: 47 Things Founders Forget

I've watched 30+ MVP launches up close. The ones that worked weren't smarter or better funded. They just shipped through a checklist. The ones that bombed shared a different pattern: a founder convinced their MVP was 'basically done' on Friday, launching Monday, and discovering on Tuesday that signup emails were going to spam, Stripe webhooks weren't firing, and analytics had been broken for two weeks.

Below are 47 items I check before any MVP I ship goes live. Not aspirational. Required. Group them by phase, run them as a hard gate, and refuse to launch with red items.

Phase 1: Pre-launch tech (15 items)

These are the items that turn into 2 a.m. Slack messages if you forget them. Most are 10-minute jobs. Skipping them is unforced error.

1. Lock production environment variables in a vault (1Password, Doppler, AWS Secrets Manager). No .env files in Git, ever.
2. Verify HTTPS works on apex and www. Auto-redirect HTTP to HTTPS. Test in incognito.
3. Configure SPF, DKIM, and DMARC records for your sending domain. Otherwise welcome emails hit spam.
4. Set up at least one error tracker (Sentry, Bugsnag, or Highlight) and confirm a test error appears.
5. Wire up a basic uptime monitor (BetterStack, UptimeRobot) hitting your /healthz route every 60 seconds.
6. Run a real Stripe (or Paddle) test purchase end to end, including the webhook that grants access.
7. Test password reset, email verification, and magic-link flows from a fresh inbox you don't own.
8. Database backups enabled and verified by restoring once. 'Backups exist' is not the same as 'backups work.'
9. Add a robots.txt and a sitemap.xml. Submit the sitemap to Google Search Console.
10. Set canonical URLs on every page. Fix duplicate trailing-slash and www variants.
11. Add a 404 and 500 page that doesn't leak stack traces.
12. Rate-limit auth endpoints (login, signup, password reset). Default 5 requests per minute per IP.
13. Audit dependencies with npm audit or Dependabot. Patch any high or critical CVEs before launch.
14. Test the full signup-to-paid flow on mobile Safari. Half your traffic will be on iPhones.
15. Run Lighthouse on three core pages. Performance score should be 70+ on mobile, ideally 90+.

Phase 2: Pre-launch operations (10 items)

Operations is what separates a launch that converts from a launch that just generates traffic. Most founders ship a working product into a broken funnel.

1. Write a one-page positioning brief: who it's for, the pain, the alternative, your wedge. Pin it.
2. Set up support email (support@yourdomain) and pipe it into a tool you'll actually check (Help Scout, Plain, even Gmail).
3. Draft a privacy policy and terms of service. Use Termly or a lawyer; do not generate from a meme template.
4. Add a cookie banner if you have any EU traffic. Cookiebot or Osano start free.
5. Write a Day 1 onboarding email that lands within 60 seconds of signup.
6. Create a Loom welcome video under 90 seconds. Embed it in the welcome email and onboarding.
7. Stand up a public roadmap (Productlane, Canny, or a Notion page) so users can see momentum.
8. Pre-write three answers for inevitable objections: pricing, security, missing feature.
9. Decide your refund policy in writing before anyone asks. 'Full refund within 14 days' is the safe default.
10. Pick the single metric you care about for the first 30 days. Activation rate is usually the right one.

Phase 3: Launch day (10 items)

Launch day is execution, not decisions. If you're still deciding pricing on launch morning, you've already lost the day. Run this list in order between 6 a.m. and 9 a.m. local time.

1. Smoke test signup, login, payment, core action, and logout in production at 6 a.m.
2. Verify analytics events are firing: signup, activation, payment. Check live in PostHog or GA4.
3. Confirm support inbox is monitored by a human (you, until you can't).
4. Post the launch tweet, LinkedIn post, and Product Hunt submission within a 30-minute window.
5. Notify your warm list (newsletter, beta users, friends) by direct email. This is 70% of day-one signups.
6. Pin a 'we just launched' banner on the homepage with a CTA to sign up.
7. Open a war-room channel in Slack or Discord. Page yourself for any error spike over 1%.
8. Reply to every comment, DM, and email within 30 minutes for the first 8 hours.
9. Keep a launch-day journal: traffic, signups, conversions, top objections. Five-minute updates every hour.
10. At 9 p.m., post a 'thanks, here's what we learned' update. This 2x's day-two engagement.

Phase 4: First 30 days post-launch (12 items)

Most founders treat launch as the finish line. It's the start. The first 30 days decide whether you have a product or a pile of code with a domain name. Treat them like a sprint, not a celebration.

1. Talk to 5 users per week. 20 calls in 30 days. No 'beta testers' replacement.
2. Track daily active users, signups, and activations on a single dashboard. Update every Monday.
3. Review error tracker every morning. Burn down the top three errors weekly.
4. Ship a small, visible improvement every 3 days. Momentum is a metric.
5. Write a public changelog. Even if no one reads it, it forces you to ship.
6. Run a churn root-cause on every cancelled customer. Email them personally.
7. Audit page-load on real-world devices weekly. Mobile speed degrades silently.
8. Set up a referral hook (in-product or simple link). Even a $20 credit drives meaningful signups.
9. Write one piece of SEO content per week targeting a real buyer query.
10. Refuse to add new features for 30 days unless 3+ users requested them. Most early features are noise.
11. Reconcile Stripe revenue with your database weekly. Webhooks fail silently.
12. Set a 30-day post-mortem in your calendar now. Three questions: what worked, what broke, what's next.

What founders skip first (and pay for)

If I had to rank the items most often skipped, the top of the list is depressingly consistent. They're all 30-minute tasks that turn into 30-hour outages.

Tools I actually use for an MVP launch

Stack opinions are cheap. Here's the boring stack I default to for launches under $5K MVR. None of these are sponsored, all of them work.

• Hosting: Vercel or Fly.io. Free tier covers most launches.
• Database: Neon or Supabase Postgres. Both have generous free tiers and proper backups.
• Auth: Clerk or Supabase Auth. Roll-your-own auth is a launch tax you don't need.
• Payments: Stripe. Paddle if you sell to EU consumers and want VAT handled.
• Email: Resend or Postmark. Mailgun if you need more volume cheaply.
• Analytics: PostHog (events + replays) or Plausible (privacy-light, simple).
• Errors: Sentry. Free tier covers most pre-revenue MVPs.
• Uptime: BetterStack. The status page is a free trust signal.

When to push the launch

Pushing a launch by 72 hours costs you nothing. Launching broken can cost you the launch entirely. If 24 hours before launch you're still missing payments, signup emails, or production analytics, push. The momentum you think you're protecting doesn't exist yet.

If you're staring at a launch date and the checklist is mostly red, that's exactly what an MVP Build Sprint is for. I run 7-14 day sprints to take an MVP from 'almost there' to launched, including the unglamorous half of this list. See the breakdown in my post on going from idea to live MVP in 30 days, or the longer take on what makes an MVP investor-ready.

FAQ