Language vs Skill App Creators in 2026, Who Fits Which

Newsthink, a YouTube channel that explains how big things get built, has run 72 paid posts for Brilliant since June 2023 in our deal log, with about 363K views per drop against a 1.21M subscriber base. Brilliant is a skill-learning app for math, science, and coding. A marketing lead at a language app messaged me Monday asking whether Babbel could buy that same Newsthink slot. The 90-second answer was no. That audience wants to build and understand things. A language app needs people who want to talk to other people. Glossary on first mention: education app (learning and language apps), CAC (customer acquisition cost), LTV (lifetime value of a customer), completion rate (what fraction of learners finish).

I sat on this post for two months because the education version of the fit question is the one operators get wrong on the first roster. The cost is not a wasted ad spend alone. The cost is a whole launch quarter spent paying for reach inside the wrong intent.

Across the deals we track, Brilliant alone runs through 572 creators and 1,983 paid posts, while Babbel runs through 129 creators and 205 posts. The two rosters barely overlap, which tells you the bookable list is split by intent long before you pick a name.

The fit question most education brands skip

What decides this is intent. Channel size matters far less. A language app sells the wish to speak with a new person. A skill app sells the itch to build or understand something.

The bottleneck is buyer intent in plain words. Most brand teams reach for follower count as the surface explanation. Babbel creators average 909K subscribers and Rosetta Stone creators average 1.03M in our log, so both language brands already skew large. Picking the biggest name does not tell you if the audience wants to learn a language at all.

Newsthink proves the split. That channel ran 72 paid Brilliant posts since June 2023 and zero language-app posts. The audience shows up to understand how a thing works. A language pitch on that channel would land soft because nobody arrived to learn French.

The four audience cuts that actually matter

We score every education creator on four cuts before a roster goes to a brand. Intent is first. Topic adjacency is second. Region is third. Cycle stance, meaning whether the creator keeps the audience through slow months, is fourth.

Intent maps to brand type. Travel, culture, and people-curiosity intent fits a language app. Build-something and understand-something intent fits a skill app. Sabine Hossenfelder, a 1.76M subscriber physics channel, ran 33 paid Brilliant posts through March 2026 because her audience arrives wanting to think harder. A language brand on that channel would pay reach rates for an audience that never wanted a language app.

The bottleneck is intent match. Reach matters far less. Artem Kirsanov, a 348K subscriber neuroscience channel, ran 21 paid Brilliant posts and pulled about 311K views per drop. That is more views than the subscriber count. The topic did the work, and the follower wall did not.

The pick your gut makes is probably wrong. Most education brands open vetting wanting the biggest learning-adjacent name they can afford. Our data says the repeat-deal pattern concentrates inside mid-size channels with one clean intent. Follower count is a weak first cut.

The creators who fit each cut

Here is how the named anchors line up against the cuts.

For skill-app intent, Newsthink and Sabine Hossenfelder both fit. Both reach audiences who arrive to learn how things work. Newsthink holds 72 paid Brilliant posts and mattbatwings, a 315K subscriber coding channel, holds 19. Brilliant keeps booking these channels because the audience already wants to build.

For topic-adjacent skill teaching, Lucie Villeneuve ran 59 paid Skillshare posts at 96K subscribers. Skillshare is a creative-skills class app. Her audience already takes classes, so the message lands at a small subscriber count. Jess Karp ran 67 paid posts across Skillshare and Squarespace on a 523K channel for the same reason.

For language-app intent, the fit pool looks different. Language brands lean on travel, culture, and people-curiosity channels. Our log shows Babbel and Rosetta Stone splitting 392 paid posts across 235 creators, and the named build-something anchors above do not appear in that pool. The past-deal log is where the real intent split lives, well before a single email goes out.

Paying for the wrong intent is the quiet roster killer.

We do the intent cut so your roster ships