--- title: Choosing creator niches by what the viewer wants (and reading the comments) slug: creator-niche-selection-by-audience-intent summary: Why on-topic creators (UFC fighters, mukbang, recipe and Love Island accounts) flopped for Cal AI while day-in-the-life fitness creators and moms printed, the comment-section test for niches and individual creators, and spotting creators before they blow up. domain: mobile-apps tags: [organic-growth, influencer-marketing, creator-selection, audience, cal-ai] related: [creator-app-fit-and-magic-moment, credibility-vs-relatability-creator-matching, influencer-deal-pricing-rpm-and-flat-rates, creator-outreach-at-scale] sources: [yt:22wFjjVyXUI, yt:l0vswWcGmMs, yt:VGJWRjz6nq8, yt:OxE79rpGioI] updated: 2026-10-24 --- # Choosing creator niches by what the viewer wants ## TL;DR - Choose niches by **why the viewer watches**, not what the creator does: UFC fighters track every calorie, but their audience is on the couch with beer and pizza — thousands of dollars, no download spike.[^1][^2] - Same failure with **mukbang** and even **recipe/cooking** content: viewers want food as entertainment or reward, not calorie bad news.[^3][^4] - **"Views can lie, followers can lie, comments cannot."** Niche test: do big creators' comment sections ask for the solution your app offers?[^5] - Creator test: "fire emoji" comments (audience there for looks) vs "thank you, I didn't know how to train hamstrings" (audience there to learn). Only the second converts.[^6] - Jake Castillo: ~**9 of 10 creators won't convert**; a cult-like following is what you want, and it's a *lagging* indicator — sign good-but-small creators cheaply before they blow up.[^5][^7] - Best Cal AI angle: big creators' **day-in-the-life vlogs** with a quick scan at breakfast, ~4 posts/month.[^8] ## Details **Failures and why (Cal AI).** - *UFC creators*: the bet was that fighters cutting weight need easy tracking; viewers enjoyed watching someone else diet, not doing it themselves.[^1][^2] - *Mukbang*: "scan 20,000 calories" seemed a wow moment, but a cheat-day viewer wants no numbers.[^3] - *Recipes, cooking, normal meals*: didn't convert; healthy-meal videos were only "okay" ([[people/alfie-dickens]]).[^4] - *Comedy/random viral creators* (a drive-thru stunt creator scanning fries): good CPMs, but the audience isn't thinking about calories.[^9] - *Love Island* contestant Harry Cooksley: decent start, then underperformed — comments were all about his partner Shakira, so the audience was there for her.[^10] **What worked.** Four fitness sub-niches: bodybuilders/gym bros, women's fitness, weight loss, plus moms. Mom influencers were expensive because they know they convert; some were profitable at a **$10 CPM**, others weren't.[^11] Big lifestyle creators (e.g. Jessie from the *Mormon Wives* reality show) sometimes drove hundreds to thousands of promo-code uses per video; Alfie timed her posts just before and after a new season to get a low effective CPM.[^8][^12] **Saturation within a niche.** One creator cost $20K for a month and drove ~$120K in sign-ups; a similar creator in the same niche did poorly because the audience had already seen it. Fitness was big enough to rotate sub-niches; smaller niches needed only one creator.[^13] **Diamonds in the rough.** Jake signed his ~12th–15th creator at 10–12K views because the content was funny and charismatic; in month three a video hit 7M views and he never dropped below 500K afterwards.[^7] **Why comments work (App Leaders, Sep 2026).** Jake credits co-founder Blake Anderson with the comment-section test, which became Cal AI's "holy grail" for predicting profitability for both influencers and UGC creators: a parasocial bond shows up as real conversations in the comments, turning a gamble into a calculated bet. Back-end revenue data from agencies can't be trusted; comments can't be faked.[^16][^17] Before signing anyone, also check the niche can pay: lots of creators, few rich competing sponsors (a $10/month app can't outbid fashion brands with ~$200 AOV), and audiences who want to become like the creator — see [[choosing-ugc-influencers-or-paid-by-stage]].[^18] **Older buyers.** [[people/dawson-gibbs]]: mom, family and 30+ creators convert highest for family-relevant products; reach older audiences on Facebook/Instagram, 18–25 on TikTok/Snapchat.[^14][^15] ## Disagreements / caveats - Cal AI's experience is one category (food/fitness). The principle — audience motivation over topical match — is the transferable part. - [[ugc-new-account-warmup]] steers *new* accounts to a niche audience; this note is about picking *existing* creators whose audience already fits. ## See also - [[credibility-vs-relatability-creator-matching]] - [[creator-outreach-at-scale]] [^1]: yt:22wFjjVyXUI @ 20:16 — Jake Castillo [^2]: yt:22wFjjVyXUI @ 21:17 — Jake Castillo [^3]: yt:22wFjjVyXUI @ 23:12 — Jake Castillo [^4]: yt:l0vswWcGmMs @ 19:19 — Alfie Dickens [^5]: yt:22wFjjVyXUI @ 25:37 — Jake Castillo [^6]: yt:22wFjjVyXUI @ 26:54 — Jake Castillo [^7]: yt:22wFjjVyXUI @ 28:41 — Jake Castillo [^8]: yt:l0vswWcGmMs @ 20:31 — Alfie Dickens [^9]: yt:l0vswWcGmMs @ 18:31 — Alfie Dickens [^10]: yt:l0vswWcGmMs @ 30:03 — Alfie Dickens [^11]: yt:l0vswWcGmMs @ 17:28 — Alfie Dickens [^12]: yt:l0vswWcGmMs @ 27:28 — Alfie Dickens [^13]: yt:l0vswWcGmMs @ 16:12 — Alfie Dickens [^14]: yt:VGJWRjz6nq8 @ 46:49 — Dawson Gibbs [^15]: yt:VGJWRjz6nq8 @ 22:45 — Dawson Gibbs [^16]: yt:OxE79rpGioI @ 19:04 — Jake Castillo [^17]: yt:OxE79rpGioI @ 20:03 — Jake Castillo [^18]: yt:OxE79rpGioI @ 21:33 — Jake Castillo