We just shipped a 33-module marketing course called Be Your Own CMO. The pitch: AI writes the copy, you understand the decisions.
Then I used it to audit the product that sells it.
The audit found four of our own pages were dead-on-arrival for the buyer we said we were targeting. The fixes took two days. Worth a write-up because the gap between "we know the framework" and "we apply the framework to ourselves" is how most marketing fails.
The product
GoaTech is the platform. We sell technical courses that teach builders how to ship production-grade software with AI. The flagship is a 33-module series — build a Shopify-quality online store from scratch. Backend, frontend, mobile, payments, analytics, growth.
The course before this audit was structured the way technical courses usually are: by what gets built. "Phase 1: Launch the Store. Phase 2: Make It Production-Grade. Phase 3: Grow the Business." Each part priced separately ($149 / $199 / $149) with a $497 bundle.
That structure made sense to me — I'm a builder. The course is built in build-order. But the buyer isn't shopping for a build order. They're shopping for an outcome.
What the audit found
Module 4 of the marketing course is Positioning. The exercise: write your one-sentence positioning statement. Audience, alternative, unique benefit.
I tried to write GoaTech's. I couldn't.
Specifically, I couldn't fit the three parts into one sentence without one of three failure modes:
- Ladder pitch: "Part 1 launches your store, then Part 2 makes it real, then Part 3 grows it." Sequential. Implies you have to buy all three. Hides the standalone value of each part.
- Bundle pitch: "Three parts that build a Shopify-quality store." Loses the per-part hook. The buyer hears "$497 commitment" before they hear "or just $149 for what you actually need today."
- Buffet pitch: "Pick whichever part matches your stage." Sounds like a hardware store. Loses the through-line.
The pages were doing all three at once. Part 1 detail page sold Part 1 in isolation. The series page (which didn't exist) was supposed to anchor the ladder. The catalog showed all three as siblings. Three different stories with no editorial control.
The four pages that were dead
1. The series landing page didn't exist.
/learn/series/vibe-northstar was a 404. If someone clicked an ad about the bundle, they landed on Part 1's detail page with no signal that Parts 2 + 3 existed. The "Save $100" pitch had no home.
2. The Part 1 detail page lacked any reference to the series.
Even after enrolling, no breadcrumb said "this is Part 1 of 3." Buyers who finished Part 1 had no surfaced path to Part 2.
3. The price anchor was upside-down.
Each part page showed a $149 / $199 / $149 number floating in space. There was no anchor against the alternative — what would this cost if I built it any other way? Without an anchor, the buyer's reference price is whatever course they bought last (probably $0 — most courses are free).
4. The "not for you if" qualifier was missing.
Module 7 of Be Your Own CMO is about disqualifiers. Telling people who shouldn't buy improves the conversion rate of who should — it raises the perceived selectivity, and it pre-empts refund requests. We had no disqualifier copy anywhere.
What we fixed
Built the series page. /learn/series/vibe-northstar now leads with a price anchor:
Building it with an agency: $170k+
A bootcamp with 1/4 the content: $15k
The full Series: $497 (three parts, take in any order, own every line.)
The agency number isn't made up — it's the real cost of an agency engagement to ship the same scope. The bootcamp number is the real cost of a 12-week bootcamp that teaches a fraction of the same material. Both are public, falsifiable references. The series price reads as the punchline, not the ask.
Added a "Part N of 3" cross-link banner on every part-detail page. Below the hero, a thin banner: "Part 1 of 3 in Launch a Real Online Store with AI · See all 3 parts →" Closes the discovery loop without forcing a redesign of the catalog.
Wrote the disqualifier:
Not for you if you've never installed a code editor, never used a terminal, or have never paid for an AI tool. Start with our free AI Fluency course first.
This costs us a tiny number of buyers who would have bought and returned. It earns us trust from the rest. Module 7 was right.
Pinned a 14-shot screenshot checklist for the campaign creative — every screenshot is a real artifact from a real working build, not a stock illustration. The course teaches "AI writes the code" and the screenshots prove the code is real.
What we didn't do (and why)
We didn't drop the price. "Free during beta" was tempting — every course platform has a $0 tier. But Module 11 of Be Your Own CMO is about pricing as a signal. A $497 course says "this teaches something hard." A $0 course says "you'll buy it but you won't finish it." We're keeping early-access pricing for now (zeroed via a feature flag) but the displayed list price is still $497, with the discount marked explicitly so the anchor survives.
We didn't add a $397 bundle SKU yet. Saving $100 across the bundle is the right move once we have data on attach rate (do Part 1 buyers also buy Part 2?). Pre-data, picking $397 vs $447 vs $447-with-cart-abandonment-discount is guessing. Post-launch we'll add the bundle SKU when there's enough data to set the price by margin instead of by feel.
We didn't write a long-form sales page. The series page is short on purpose — five sections, no scroll-walls of testimonials we don't have yet. Module 6 of Be Your Own CMO covers "the page that fits in a screenshot." Our series page does. Adding length would dilute it.
What this audit is not
It's not a brag. Every founder finds their own product is broken when they look at it through fresh eyes. The lesson isn't "I'm clever for finding these gaps." It's "writing a course about a thing forces you to apply the thing, and that finds gaps that years of staring at the page didn't."
It's not a finished list. There are at least four other audit findings I haven't gotten to yet — the catalog sort order, the email sequence after enrollment, the post-purchase upsell, the Spanish edition. Each will be its own post when we ship the fix.
The course that did this
The course is called Be Your Own CMO. It's 33 modules (the same shape as the technical courses we sell). Audience: solo founders who can't afford an agency and don't want a coordinator role.
If you're shipping a product right now and your homepage doesn't pass Module 4, you'll know inside 20 minutes.
If you read this and shipped a fix, tell me what broke. The next post in this series is about the four findings I haven't fixed yet.