Source
rawraw/how-i-keep-my-app-stable.md
url

TL;DR: Chris Raroque protects Ellie with “spring cleaning” blocks after major features: spend two to three weeks clearing small bugs, user requests, optional preferences, and integration improvements that make the app feel stable without becoming a new flagship project.

Biggest lessons

  • Major features create a cleanup backlog. While Chris focuses on big work like widgets or integrations, smaller requests pile up. He deliberately blocks time after the push to address them.
  • Small bugs can be simple but consequential. One Apple Calendar bug was just a control-flow issue: Apple events only loaded after Google events, so users with no Google events never saw Apple events.
  • Optional settings can resolve taste conflicts. A requested iOS layout change made estimate/actual chips more prominent, but Chris shipped it off by default behind a setting because not everyone would want the desktop-style density.
  • Polish compounds. Quick-add shortcuts, Zapier directionality, settings labels, help articles, and small layout work are not flashy, but they make the product feel maintained.
  • Feedback boards turn cleanup into a system. The work queue comes from user-visible demand, not only the founder’s guesses.

Why it matters

  • This source extends app-product-craft with a maintenance cadence: product quality is a scheduled practice, not an occasional emergency.
  • It connects to product-led-growth because a self-serve product has to keep earning trust after activation.
  • It also supports launch-playbook: feedback and analytics infrastructure only matter if the team actually revisits what users report.