Home • About • Projects • Blog • Resources • Contact
Five days, countless commits, and some of the most foundational changes to Launchling yet.
Since my last update, Launchling has grown up fast. What started as a scrappy MVP now feels more like a product with an identity; visually, structurally, and technically. Here's what’s new and what I’ve learned:
🛠️ 1. Laying the Groundwork: Routing, Pages & Layouts
I did a full routing overhaul to lay proper foundations:the homepage (/) is now a true landing page; the onboarding form lives at /start ; and I've added new pages for /about, /how-it-works, and /privacy. These use a shared layout component with a consistent navbar and footer; with print views excluded so exported plans stay clean.
The nav bar and most of the components are now mobile responsive, with a hamburger menu and smoother UX all around (this was quite tricky as I’ve never been too strong with CSS). These might seem like small wins, but the traffic I have gotten so far is largely from mobile, so it felt fairly urgent.
💫 2. Hero Blocks & Plan Preview
To keep things modular and re-usable, I introduced a HeroBlock component that now powers all hero sections from the landing page through to the thank you page.
More importantly, Launchling now shows a real plan preview on the homepage via a PlanPreviewMobile component; giving users a tangible peek at what they'll get, and hopefully giving them the confidence that it’ll be worth their while to spend a few minutes filling out the form.
📥 3. Email: From Broken to Bulletproof
This part took days. For ages, Launchling’s emails simply… didn’t work. To begin with I couldn’t get them to send at all, when finally I fixed this, they arrived hours late. The culprit turned out to be a combination of broken setup, DNS misconfigurations, and provider limitations.
I tore it down and rebuilt it: switched away from Resend, configured a custom domain email (hello@launchling.io), and finally wrangled DNS, SPF, and DKIM config into shape.
It was an absolute pain, but it was worth it as it’s pretty solid now.
🧠 4. New Feature: The Risk Mapper
I built a brand new feature called the Risk Mapper. This is a really simple tool to help early-stage founders think through what might make their idea risky and how to tackle it; as well as what map the assumptions they’re making that will need validating.
It's beginner-friendly, editable, and linked to the user’s plan. I’m also hoping it’ll be a key hook: users can unlock it by giving quick feedback and an email address which helps me validate the product and gives me better signal on what’s working (and what’s not)
It's also the first step toward layering in smarter, more contextual guidance after the initial plan is delivered.
📱 5. Fixing Form Frustrations: A UX Overhaul
The onboarding flow, previously a rubbish pants on mobile, is now smoother and more accessible, and generally less prototypey.
Major improvements include:
- Generally improved styling and responsive layout
- A step indicator and animated transitions
- Accessible focus states, improved validation with visible error states
I also created a SkillRatingGroup component to keep input steps consistent and tidy.
🧪 6. Code Quality and Test Coverage
Launchling now has real test coverage, with 110 passing tests across 22 test suites. As a former engineer who used TDD (test driven development), this felt like a minimum especially as the complexity grows, it’s how I’ll continue to move fast without breaking things.
🔍 Current Test Coverage:
- Overall: 80.53%
- Utils: 100%
- Components: 79.7% (with several at 100%)
- Pages: 73.5%
- Branch coverage: 71.46% (still room to grow)
Areas like RiskMapper.tsx and PlanContent.tsx have lower coverage and are now on my hit list for improvement. But overall, this is a solid foundation to build from.
Every component is now more modular, better typed, and accessibility-conscious. Tests pass, code is clean, and the whole thing feels easier to extend.
What’s Next?
Now that things are stable, I’m shifting focus toward growth and learning:
- A lightweight idea check to improve prompt quality (and reduce GPT costs)
- A review of the question set to ensure we’re only asking questions that deliver value
- Plan confidence scores
- Private plan pages with follow-up triggers
I’m hoping to soft-launch soon and test if there’s a market here: people who want to start, but feel too unsure, alone, or overwhelmed to try.
—
Try it → launchling.io
Give feedback to unlock the Risk Mapper
Or subscribe to hear when the next update drops.