Mobbin

The channel is highly learning-useful, pairing large analyzed samples with concrete frameworks and acknowledged limitations.

• Specificity: descriptions name exact sample sizes, named case studies (Duolingo, Shopify, Mercury), named researchers (Wolfram Schultz, Lally et al.), and topic-level chapters.
• Actionable detail: episodes cover test frameworks, conversion metrics, compliance caveats, and practical alternatives like pause options and exit-survey design.
• Limit: curated metadata cannot confirm execution quality of the analyses; one comment flags a discrepancy between a claimed 1,460 flows and Mobbin's apparent 935-filter count.
• Honesty signal: the cancel-flow episode explicitly discloses iOS under-representation in its data.

Curated content leans on stated, quantified studies with disclosed caveats, and comments mostly add verification rather than contradiction.

• Evidenced claims: repeated large-sample study framing plus explicit caveats (iOS Settings cancellations, B2C-only onboarding data) support credibility.
• Correction pressure: commenters challenge causation (long onboarding vs. success), question sample figures, and note screenshot patterns lack conversion evidence, indicating claims are falsifiable and discussed.
• Limit: comments note the Mobbin MCP video produced mixed real-world results for paid users and functions partly as promotion, slightly lowering confidence for that episode.

Most episodes are standalone research analyses, but one video is explicitly sponsor-aligned with Mobbin.

• Signal: the dashboard and onboarding studies link to free Mobbin blog write-ups; interview content monetizes indirectly rather than pushing paid offers.
• Signal: commenters repeatedly flagged the Claude/Mobbin MCP video as effectively an advertisement, citing quarterly-minimum pricing and no trial.
• Limit: no evidence of coaching funnels, paid courses, or affiliate pitching in the remaining curated descriptions.

Each upload is a distinct original study or interview rather than recycled material, with some deliberate companion pieces rather than duplicates.

• Signal: sample sizes, topics, and chapter structures differ across all eight videos; no near-duplicate uploads appear in the curated set.
• Signal: the 4,700-paywall episode is labeled a companion to a prior paywall episode while covering distinct material (early optimization, metrics timing).
• Limit: recurring Mobbin sourcing and shared paywall/onboarding subject matter create thematic overlap across episodes.

The sampled comments show substantive, critical, practitioner-level discussion across most episodes.

• Critical engagement: commenters challenge causation in onboarding findings, question methodology (paywall variant exposure, filter counts), and debate whether streaks undermine intrinsic motivation.
• First-hand experience: replies include a Lyft onboarding designer's perspective, creators' abandonment data from their own flows, and reports of MCP output falling short of promises.
• Constructive exchange: multi-reply threads resolve questions (Figma connector setup, alternative billing rules) and propose follow-ups (Adobe cancellation, B2B onboarding).
• Limit: evidence is a newest-first selected sample with no like counts, identities, or totals, so breadth and representativeness of overall discussion cannot be judged; some comments are simple tool questions or requests for text versions.

Newest video 11 days ago · 10 uploads in the last 12 months