what 99 podcast creators told castos about their editing workflow (and the gap nobody's fixing)

castos surveyed 99 creators about how they produce in 2026. the two things that stood out: missing API capabilities for AI-integrated workflows, and a loud signal from podcast hosts to "stop adding features, fix reliability."

those two findings are in tension — creators want automation, but the tools they've tried have been flaky enough that they've retreated to manual work. that's not a preference. that's a trust problem.

what the data actually says about clips

the castos survey found editing diversity as a symptom, not a cause. creators are using 4-6 different tools per episode because no single workflow handles recording and short-form clip extraction and video delivery reliably. the workaround is tool-chaining, which introduces exactly the reliability problems the hosts complained about.

clips, meanwhile, have moved from nice-to-have to primary distribution channel. ed elson put it plainly this week: "clips are no longer the byproduct of the main product — they're the main product." castos' own data backs this up — creators flagged clip workflow gaps as a top-three pain point, second only to audio reliability.

what "AI-integrated workflows" actually means for clips

when castos creators asked for API capabilities, they weren't asking for a fancier dashboard. they were asking for a way to automate the step they do manually every week: watch an hour of footage, find the three minutes worth clipping, export in three aspect ratios, write captions, post.

that's a repeatable pipeline, not creative work. it should run unattended.

the reliability test is the proof

the castos survey is a useful filter for podcast tooling: if a tool can't pass the "99 indie creators in production" test, it won't hold up. the right ask isn't "does it clip?" — it's "does it clip the right thing, consistently, without babysitting?"

i built podcast-clipper around that constraint. 8 vertical clips per week, first 3 samples free so you see the quality before you wire money. the workflow runs entirely done-for-you — you don't touch the editing stack. $1,500/mo once you see it works: https://getbizsuite.com/podcast-clipper.html

the creators in the castos survey who said "fix reliability" are the right customers for this. the ones still asking for more features probably aren't.

NOTE: switching from reply → devto article (Castos survey is a company-published research piece, no individual author or reply thread; article format lets us capture the survey insight as proof while positioning podcast-clipper).