Source: Product Strategy
File: PRODUCT_STRATEGY.md in llc-course root
Type: Internal strategy document (April 2026)
Explains the strategic decisions behind the dual-product structure — audience targeting, free-plus-paid model, pricing, distribution, and deliberate non-decisions.
Product family
| Product | Price | What it is |
|---|---|---|
| Web app | Free | 10-module interactive course on Cloudflare Pages |
| Companion PDF | 38 | Paid PDF — exercises, annotated solutions, depth (3 tiers) |
| Exercises repo | Free | VHCosta/c-systems-lab-exercises — starter files |
Target audience
Primary: OOP developers (Java, C#, JavaScript, Python) who already write software and want to understand what happens underneath their abstractions. Not beginners. Not embedded engineers already fluent in C.
Why this framing matters: Most C courses treat the reader as a beginner. This course starts at “you already know how software works — here’s what C actually does, and why.” The audience’s existing mental model (GC, references, boxing) is used as a scaffold, not an obstacle.
Key strategic decisions
Free web app
- Trust before the transaction — a user who has completed three modules knows the material is serious; they are the right PDF buyer
- Discovery without friction — shareable on HN/Reddit without paywall creating awkwardness
- Self-selection — free users who reach Module 3 have already invested time and demonstrated interest
- The free product stays genuinely free — no freemium gates, no email capture
Why a PDF, not extended web content
- Exercises with real feedback require a compiler; a simulated terminal cannot provide this
- Annotated solutions explain not just the correct answer but why the constraint holds
- The “why” layer — what the standard requires, what the optimizer assumes, what production codebases do
- Reference material (debugging appendix) is better as a PDF
Why Typst, not LaTeX
Sub-second builds, clean code blocks, no package hell. Trade-off was speed and simplicity vs. ecosystem maturity — Typst wins for solo maintenance.
Why Gumroad, not Udemy/Coursera
Gumroad takes ~10%; Udemy takes 50–75%. Udemy’s discount culture attracts the wrong buyer. Gumroad allows full brand control and no platform dependency.
Why not video
Production cost, searchability, audience preference (experienced developers read faster than they watch), and maintenance — updating a chapter in Typst takes minutes vs. hours for re-recording.
Pricing rationale
| Tier | Price | Contents |
|---|---|---|
| $18 | All 11 chapters + exercises, no solutions | |
| PDF + Solutions | $28 | Above + annotated solutions (marked “Most Popular”) |
| Complete | $38 | Above + extended projects |
- $18 — below the “I should think about this” threshold, impulsive-buy territory
- **40–$60 for PDFs
- **28 to not feel dramatic; extended projects are substantial work
Three tiers vs. two: anchor buyers to the middle option while giving a clear lower and upper path.
Deliberate non-decisions
Community (Discord/Slack), cohort runs, certificates, interactive code execution, mobile app, video course, beginner track, paid web content. Each explicitly rejected with a stated reason in the document.
Notable: Interactive code execution is explicitly called out as not being built (“requires a backend, sandboxed execution environment, and ongoing infrastructure costs”). This is now being revisited — see meta-terminal-plan.
Open strategic questions (from the document)
- Game Boy emulator as PDF ch12 — strongest upsell argument; see meta-gb-emulator for full analysis
- PDF versioning / buyer updates — free updates via Gumroad file update; price bump if ch12 added for new buyers?
- Long-term pricing — launch prices are intentionally low; room to raise as PDF reputation builds
- Exercises repo as independent product — currently a delivery mechanism; could become a discovery channel with better README/topics
- Expansion beyond C — Rust systems course or OS internals as a second product; separate brand question