Pick a niche and positioning, recycle code, databases and components across products, and launch weekly; judge each by paid revenue, not downloads or active users.
When it applies: pre-PMF founders with reusable stack
Each week build something small, launch it on every channel, then post the result, and repeat. Each finished project is a dice roll and a fresh reason to post; don't clone what worked, twist it so it's novel.
When it applies: Early-stage indie hackers without a hit yet
He never spends more than about a month on an unproven product; if the launch brings a little revenue but no excitement he keeps it live (no shutdown) but stops investing time.
Run 'one product per week until something sticks' so you ship instead of plan, get feedback fast and drop weak ideas; 10 ideas at 10% should yield one winner.
When it applies: founders without deep industry insight
If a project gets a few users and a little money after launch, resist turning it into your identity; launching new things is faster and each launch grows your audience and learning.
After two weeks ask: are people signing up without chasing, coming back on their own, paying without heavy convincing, and giving feedback? Yes to all four: continue. No to any: kill it unless the fix is obvious.
When it applies: early validation of each new product