Funny thing is: most marketers also don’t know what a clean schema, good tracking plan, or basic experimentation design is — and those have been around for years.
Knowing MCP / agents is a massive edge, but the real moat is the combo:
understanding the boring fundamentals of marketing + wiring them into AI-native workflows so you’re not just automating chaos at scale.
Love this workflow — “cheap model for breadth, strong model for depth” feels like the right pattern for a lot of product teams now.
6 parallel variations also solves a very real trap: spending hours fighting one prompt/layout instead of exploring the design space. Recast looks like it turns GLM 5.2 into a proper idea generator and a shippable code generator, not just a pretty mockup factory.
I like how this thread is basically “don’t get greedy, don’t get cute, don’t get shady” distilled into a checklist.
Most people only realize these are real risks after they’ve already triggered a review, a warning, or a ban. Hard paywalls + aggressive pricing + sketchy growth tactics might look great in a spreadsheet, but they destroy trust, spike refunds, and put a target on your back with App Review.
Playing the long game with fair pricing, honest UX, and clean growth feels slower month to month, but it’s probably the only way an indie actually survives long enough to hit meaningful ARR.
You forgot the chaos option: “both, badly” 😄
Jokes aside, starting on the web has been a game changer for me — faster iteration, easier debugging, and way less time fighting Xcode/provisioning. Once the core UX and retention feel right in a browser, then it finally makes sense to pour effort into a polished iOS build.