First month with your boilerplate: a real roadmap for indie hackers

Buying a boilerplate and shipping with it are different things. Some open it on Saturday, spend 2 hours, close it "to study it properly later," and never touch it again. Others go full speed, burn out in 4 days, and quit at the first friction point. This guide is the roadmap for your first 30 days with the boilerplate. Aimed at indie hackers with ~10h/week of real time (not the 40h you promise yourself on Sunday). ## Week 1: setup and first contact **Goal**: boilerplate running locally with your name and your landing. - Day 1 (2h): clone, install, set `.env`, run `bun dev`. Confirm everything works. - Day 2 (2h): change logo, colors, domain. Rewrite landing copy for the ICP you're validating. - Day 3 (1h): customize emails (welcome, verify) with your copy. - Day 4 (2h): read the root `CLAUDE.md` and the per-feature ones you'll touch. - Day 5 (2h): explore the admin dashboard, signup flow, billing flow. Buy from yourself as a test. **Week milestone**: if a friend signs up and finishes onboarding, it works end-to-end. ## Week 2: your first differentiating feature This is where you stop being "another SaaS with auth" and start having a product. - Day 1 (2h): define on one page what your product does. Data model. Endpoints. - Days 2-3 (4h): ask Claude Code to implement the Prisma model + server actions. Follow the boilerplate rules. - Day 4 (2h): basic UI for that feature. - Day 5 (2h): manual testing. Obvious cases (what happens if the form is empty?). **Week milestone**: your feature works locally. Not pretty, but it works. ## Week 3: polish and production - Day 1 (2h): improve the feature's UI. Loading, error, empty states. - Day 2 (2h): connect the real payment provider. Configure products. Webhook to localhost with a tunnel. - Day 3 (2h): provision VPS (Hetzner) and install Coolify. - Day 4 (2h): first deploy to production. Configure domain + SSL. - Day 5 (2h): real payment test with your own card. Verify the role flips. **Week milestone**: you're in production, someone could actually buy. ## Week 4: real launch - Day 1 (2h): email your list (even if it's 30 people). Explain what it does and for whom. - Day 2 (2h): LinkedIn post. Honest tone, no fake marketing. - Day 3 (1h): post in IndieHackers / a Discord where your ICP hangs out. - Day 4 (2h): enable analytics (Plausible, Umami, Clarity). Start measuring. - Day 5 (2h): talk to the first 3-5 people who sign up. Ask what's missing. **Week milestone**: you have the first 5 real users and conversations with them. ## What NOT to do in the first month **1. Change stacks**: if you bought the boilerplate and mid-way think "I'd rather use Astro," you abandon and start over. Suicide. **2. Build unvalidated features**: "what if I add a referral system…" before having 5 customers wastes time. **3. Rewrite the boilerplate's code**: you'll be tempted to "clean it up". Rewrite ONLY if it blocks you from shipping. Otherwise, live with it. **4. Wait for 100% polish to launch**: you launch on week 4 Friday. Ugly bits get fixed in production with real feedback. **5. Not measure anything**: starting without analytics is flying blind. 30 minutes of Plausible/Clarity save you later. ## If you're not in production after 30 days Something's wrong. Options: - You underestimated the boilerplate curve → cut scope, keep the plan - You picked a feature too complex → simplify to MVP-of-MVP - The boilerplate doesn't fit → you found out with data, decide - Procrastination → fix the system, not the plan I always end up talking to someone who's been "getting ready to launch" for 3 months. The problem is almost never technical. It's the system, not the plan. ## Bottom line The first month with a boilerplate isn't measured in lines of code. It's measured in whether you have something in production to send people to. If you haven't launched by day 30, it's not the UI components. It's the method. Cut scope and push. Polish comes later, with customers and feedback. Before that, there's nothing to polish.