Post-Launch Polish: What I’ve Been Improving on TinyLaunch

HomeAboutProjectsBlog Resources Contact

Update: TinyLaunch is now Launchling! Same idea, new name and better experience.

image

A few days ago, I quietly shared TinyLaunch — a lightweight tool that helps early-stage, non-technical founders get from “vague idea” to clear next steps.

No formal launch. Just a few messages to people I thought might find it useful.

But even a trickle of early usage was enough to highlight what needed fixing — especially on mobile.

This post is a behind-the-scenes look at what I improved, what broke, and how I’m using tools like Cursor and ChatGPT together to build and ship faster.

🛠️ Fixing the Experience: The First 24 Hours After ‘Launch’

Even though the MVP was “functional,” a lot of small UX issues became glaring once real users started exploring.

🧵 Accordion Overhaul

Plans were being presented as giant walls of text — fine on desktop, overwhelming on mobile. I refactored the output into an interactive accordion component:

  • Collapsible sections
  • "Expand All" / "Collapse All" toggle
  • Highlighting key sections (e.g., Tools, Risks, Milestones)
  • Smooth animations + localStorage persistence

This instantly made the plans more digestible.

🎯 Visual Hierarchy for Feedback

One key flaw: the feedback form looked like part of the plan. I redesigned it as a clearly distinct card with a bold call to action and better spacing — especially on mobile.

🧪 Writing Tests (Yes, Even for a Tiny MVP)

Having moved over from no-code tooling to a fullstack codebase, I started to see that changes were becoming tougher, so I asked Cursor to scaffold unit + integration tests across:

  • Utilities like validation + rate limiting
  • API endpoints for plan and feedback submission
  • Core components like MultiStepForm and PlanAccordion

I hit common issues: ESM vs CJS, CSS module stubs, Jest matchers. Cursor did a great job resolving 90% of them, and now I’ve got ~95% test coverage across core flows. I think this will make a huge difference going forward.

🧠 Using ChatGPT + Cursor Together

My workflow now looks like this:

  • ChatGPT helps me think through UX decisions, phrasing, and architectural tradeoffs
  • Cursor helps me implement and debug with full project awareness

I rarely use either in isolation. Together, they’re my current AI pair programming stack.

📱 Mobile View: From Afterthought to Priority

Post-launch, I realised how many early users were mobile-first. So I:

  • Rewrote all layout containers to behave well on narrow screens
  • Restyled form steps for readability and padding
  • Refactored radio/checkbox inputs for accessibility

And I fixed bugs like:

  • Inputs overflowing containers
  • Weird shadowing and inconsistent toggle buttons
  • Feedback elements being indistinguishable from the content above them

🖨️ New: Print View

For users who want a copy of their plan:

  • I added a “Print-Friendly View” that expands all sections and strips interactivity
  • It’s clean, readable, and shareable — which also helps with virality

🧪 What’s Next

Here’s what might be coming soon:

  • Plan confidence scoring based on feedback data
  • Smart validation that checks plan alignment with user inputs
  • A more dynamic homepage based on use cases
  • Optional AI walkthrough of your plan

💬 Want to Help?

This project is still tiny, but tiny ideas become real when you build, test, and learn (which is kind of the whole point of TinyLaunch).

Every plan now includes a clear feedback section. If TinyLaunch helped you, or didn’t, please take the time to fill this in or drop me an email with any questions or feedback - hello@emilysas.com

🪂 Try it now

Turn your idea into a tiny, tangible prototype — with a little help from Launchling.

👉 Try Launchling – no signup, no jargon, just a gentle push forward.