What we'd rebuild in Teachable's funnel

We audited Teachable's public funnel the way we'd audit a client's, mapped onto the eight stages of the Growth System, with one honest constraint: we can only see what any visitor can see. No conversion rates, no traffic mix, no in-app flows. So this teardown makes zero claims about internal metrics; the bottleneck ranking below is modeled leverage, not measured impact. Inside data could reorder it; that's true of every outside-in read, including the ones you've seen from anyone else.

A teardown that only criticizes is a sales pitch in a lab coat. The strengths are real: processing fees, chargeback fees and per-plan caps are all published; many competitors hide these. The changelog shows substantive releases roughly every two weeks. Social proof is specific. “100M+ students”, “180 countries”, a SOC 2 badge. The blog content is genuinely good: TL;DRs, comparison tables, implementation checklists. And the migration offer for established course businesses is a smart, well-targeted switcher play. The leaks below are presentation and architecture problems, not honesty problems.

Every visitor, from every channel, passes through one sequence, and it fights itself four times. The homepage says “Start for free”; the pricing page says “7-day free trial. 30-day guarantee.” There is no permanent free plan. The $39 Starter plan carries a 7.5% transaction fee; the fee drops to 0% only at $89. That is the single most decision-relevant fact on the page, and it lives in table fine print. Student caps come with a conditional rule (“once you reach $10k USD/year in sales… these limits are removed”) stacked onto four creator plans and an enterprise tier. And every “Start for free” button deep-links to a signup URL preselecting the lowest-tier monthly plan, the one with the 7.5% fee, regardless of what the visitor was just reading.

This ranks first because 100% of traffic passes through it, and because it's the cheapest possible fix: copy and routing, no product changes.

The homepage sells to solo creators (“Whatever your expertise”, with examples from fitness to coding). The blog's recent posts court enterprise L&D buyers (“Best employee training tracking software”, ending in “Request an Enterprise demo”). The newsletter brand says “Join the 9 to 5 Quitters Club.” Three different customers, one landing surface, one CTA. Whatever traffic the SEO engine wins, the message mismatch spends. It's the content-to-landing version of the classic ad-to-landing leak.

The rebuild: segmented landing paths. Enterprise content lands on an enterprise page; creator content keeps the creator homepage; the newsletter splits into tracks whose promises don't contradict the enterprise pitch. This is systems work, not creative work; the assets already exist.

Between “read a blog post” and “book a demo / start a trial” there's no visible middle path on the surfaces we fetched. Checklists and calculators sit ungated; the only newsletter signup hides in a dropdown and promises no concrete value. For a considered purchase like a platform migration, most readers aren't demo-ready on first touch. Today they leave and nothing is owned. (Fairness note: a JS popup our fetcher can't render may exist; the visible architecture still routes readers to a binary choice.)

The rebuild: gate the genuinely good assets that already exist (the checklists, the comparison frameworks) into segment-specific lead magnets feeding automated nurture tracks. This is the CRM and Automation layer of the Growth System doing its actual job: catching demand that isn't ready yet.

The signup flow itself returned 403 to our fetcher, so we make no claims about form fields, step count, or whether a card is required for the trial. We didn't audit paid ads, JS-rendered popups, or anything behind the auth wall. Teachable is not a client and hasn't asked for this; it's our outside-in read, published because it shows how we think. If someone at Teachable spots an error, tell us and we'll correct it and say so.