In the same vein as the 10,000 hour rule, I will claim that it takes 10,000 edit/refresh cycles for one to get to an excellent product. From the first prototype that you stand up, through lots of polishing and tweaking, making the user experience delightful to shaving critical milliseconds of unwanted latency, there’s a lot of repetitive editing and refreshing.
Everything that stands in between your edit to your refresh is slowing down your product honing process. Be it compilation, reinterpretation, writing unit and integration tests, or getting approvals from peers / legal – these will all either set you back time-wise or will prevent you from achieving 10,000 edit/refreshes and degrading your product. So make sure you’re unblocking your critical path on the product. Add tests and optimization steps cautiously – they should come in later and you should have a good idea of what reward you’re getting.
P.S. I’m pretty sure there’s a Tim Ferris of product that will find ways of short-cutting the 10k process. I’m also sure some tools and methodologies are totally worth it despire upfront cost. As always, take this post as a mental challenge more than a prescribed way of living.