You should do manual work for your early customers, as it's crucial to get those initial users, and the experience will provide valuable insights for your startup's development.

When sending cold emails, do it manually at first to learn what works before trying to automate the process.

It's acceptable to be inefficient at level one as long as you deliver exceptional customer satisfaction, which can sometimes mean manual processes.

The main goal of doing things that don't scale is to maximize learning and delight early customers, even if it requires manual and one-off efforts.

Learning by doing is essential; producing designs, even bad ones, helps you understand what works and what doesn't.

Doing things that don't scale means engaging in handmade, artisanal efforts early on, even if you know you can't maintain them forever, because it helps you learn and grow your startup.

You initially have to do things that don't scale in order to get into scaling, like manually getting users or doing things that wouldn't work with a large user base.

When starting a business, you need to experiment and see what works before analyzing the results.

It's important to quickly build and test ideas. If something doesn't work, it's better to kill it and start again rather than justifying its existence with additional features that may not address user needs.