Based on research:

1. Time-box your support obligation, explicitly, in your terms. TWN caps support at 2 years post-purchase. As the only support person you have, an open-ended promise is your biggest hidden liability — write a limit down now, even generously, rather than never.

2. Lesson-level discussion threads, not a global forum. Kajabi's data: completion jumps from ~42.6% to ~65.5% when Q&A lives on the specific lesson. Answer once, serve every future student.

3. Ship a project + feedback loop, not just video. Skillshare's core insight: video + submitted artifact + your feedback beats passive watching. Keep it async so you can batch-review.

4. Use your existing content (blog, YouTube) as the actual acquisition engine — not a side project. TWN's entire funnel is 1.4M YouTube subscribers; KodeKloud's is free labs. You already have FullStackFusions — treat it as tier-1 infrastructure, not marketing fluff.

5. Fewer, deeper flagship offerings beat a sprawling catalog. TWN sells ~2-3 high-ticket programs; KodeKloud sells dozens. As a solo creator, depth on 2-3 courses is more sustainable than breadth across 20.

6. One-time payment, high perceived value, is a legitimate model — you don't need a subscription. TWN proves a structured, outcome-oriented one-time program can command a premium without recurring-billing complexity you'd have to build and support solo.

7. Decouple course from price (Products vs. Offers). Same course, multiple wrappers — one-time, payment plan, bundle. Model this in your schema now even if you only use one offer today.

8. Automate the first five minutes post-purchase. A short onboarding sequence (welcome, "start here," what to expect) cuts "how do I even begin" tickets before they happen.

9. If your content is technical, remove environment-setup friction. KodeKloud's in-browser labs turn "watching" into "doing" without asking students to configure anything locally — even a lightweight version (pre-built sandbox link, Docker Compose starter) captures most of this benefit.

10. Frame courses around a concrete, checkable outcome, not just topics covered. KodeKloud leans on real certifications; you don't need a third-party cert to borrow the trust-building mechanic — "by the end you'll be able to X" does similar work.

11. A small, free, no-signup-friction product is worth building as a funnel. KodeKloud's 100-Days challenges and Playgrounds require no credit card — the free tier does the convincing before the paid tier does the selling.

12. Track attribution per sale. Tag every Stripe transaction with source (blog, YouTube, direct). Costs nothing to add now, tells you later which content actually converts.

13. Consider cohort-style pacing over pure self-paced, at least optionally. Concentrates your Q&A into predictable windows instead of a constant trickle from students scattered across every stage — TWN avoids this by not enforcing it, but explicitly warns students that skipping ahead breaks later modules, i.e., sequence still matters even without a hard deadline.

14. Specific, outcome-quantified testimonials outperform generic praise. TWN's "10+ interviews, 4 offers, hired in London" beats "great course, would recommend" — ask future students for the specific before/after, not a rating.

15. You already avoided both Udemy's and the subscription-lock-in trap. Owning your stack means no one can cut your revenue share, and a one-time-payment model (validated by TWN) means you're not obligated to keep adding content forever just to justify a recurring charge.