Strategy & Atlas Lead
Translates data into strategy and operates the Oura Atlas loop.
EdTech: 40 courses live — ChatGPT recommends competitors, not us
An online course platform with a large catalog but thin course pages. The head of curriculum built answer-first learning pages worthy of AI shortlists.
The buying question this story answers We have a full course catalog — why does AI recommend competitors instead of us?
The gist
In EdTech, a large catalog without citable course pages means invisible in AI recommendation. Syllabus is SEO.
EdTech / Online education
Anonymized client
“New students said: I asked ChatGPT for the best data analyst course in Indonesia — we weren't named. We have 40 courses.”
Full story
Enrollment retrospective: discounts up, growth slowing
The head of curriculum sat in the quarterly retrospective. Enrollment still rose — but almost entirely from 70% discounts and affiliates. Margin per student fell. Churn rose: students complained courses “weren’t what they expected.”
New student onboarding survey: “I asked ChatGPT for the best data analyst course in Indonesia — your name didn’t come up.”
The platform had 40 courses. LMS material was strong. Public pages? Posters.
Curriculum POV: great material, bad storefront
As head of curriculum, my frustration was specific: we invested hundreds of hours per course — modules, quizzes, projects. But public course pages were 200 words of marketing without syllabus, measurable outcomes, or FAQ on beginner fit.
AI cannot cite what it cannot extract. Competitors with tidy syllabus pages? Recommended. Us? Not named.
What they tried
15 new courses per quarter. Influencer affiliates. Marketplace discounts. Broad Facebook interest ads.
Enrollment spiked temporarily. Organic and AI recommendations did not move. Churn rose from expectation mismatch.
Audit findings: 34 courses invisible to machines
6 of 40 had citation-worthy structure — rest isolated without learning paths.
No Course schema — duration, level, credential unreadable by engines.
Skill recommendation queries — competitors had comparison + FAQ; we had “enroll now” CTAs.
Churn correlation — students from thin pages churned 2× vs those who read full syllabus (internal data).
Curriculum decision: pause 15 new courses, rewrite 12 priority pages. Growth marketing objected. Data won: 12 answer-first pages projected more organic enrollment than 15 unstructured new courses.
What we ran (Oura Content — curriculum lens)
- Prioritize 12 courses — enrollment + search intent, not alphabetical.
- Answer-first course pages — audience, prerequisites, syllabus, outcomes, FAQ.
- Course schema + learning path pillars — skill clusters.
- Comparison pages — honest, citation-worthy.
- AI-referral → enrollment — separate GA4 channel.
- Curriculum sign-off — syllabus accuracy before publish.
How to read the numbers below
Typical patterns for mid-size EdTech platforms. Content engagement patterns:
- AI shortlists usually appear after 10–12 structured course pages + FAQ — months 3–4.
- Enrollment from AI-referral converts higher than discount traffic — expectations already matched.
- Churn drops 8–12 weeks after transparent syllabus — student success metric.
EdTech is a category where public syllabus = sales page worthy of citation — not just the LMS gate after payment.
Before Oura
What they had already tried
- Added 15 new courses per quarter — new enrollment did not offset churn on old courses
- 200-word course pages + video embed — no structured syllabus, duration, or outcomes
- Influencer affiliates — enrollment spikes, organic flat
- 70% discounts on course marketplaces — margin down, brand value weakened
Turning point
Audit findings that changed direction
- 40 courses — 6 pages with structured syllabus, rest marketing descriptions
- Query "best [skill] course Indonesia" — competitors named by AI, platform not
- No Course + FAQ schema — engines could not extract duration, level, certificate
- Learning paths not connected — each course an island, no skill pillars
Before & after
What changed when the strategy ran.
Before
- 40 courses · 6 structured pages
- 0 AI recommendations for priority skills
- Enrollment growth slowing · high churn
- 80% acquisition from paid + discounts
After Oura Content (target)
- 12 answer-first course pages + 3 learning path pillars
- Named in AI shortlists for 4 skill queries
- Organic + AI enrollment +15–25%/quarter (target)
- Churn down after expectations match syllabus
The challenge
Challenge
EdTech platform with 40 live courses, 12,000 active students, enrollment growth slowing. Average course page: 200 words of marketing — no structured syllabus, learning outcomes, duration, or level comparison. When prospects asked ChatGPT "best data analyst course in Indonesia" or "learn Python for beginners, which platform", competitors appeared — this platform did not. Affiliates and discounts raised enrollment temporarily; organic flat. High churn on older courses because expectations did not match thin descriptions.
The approach
Approach
- 01
Audit 40 courses — prioritize 12 high-enrollment + high-search-intent skills.
- 02
Rewrite answer-first course pages: who it's for, prerequisites, syllabus, outcomes, duration, certificate.
- 03
Course + FAQ + learningResource schema — extractable for AI.
- 04
Learning path pillars: Data Analyst, Digital Marketing, etc. — internal link clusters.
- 05
Comparison pages: platform vs bootcamp vs self-taught — honest, citation-worthy.
- 06
GA4 AI-referral → enrollment — measure self-research channel.
Execution phases
How the work unfolded month by month.
-
Month 1
Course & intent audit
Prioritize 12 courses from enrollment + search intent. Benchmark competitor course pages cited by AI.
-
Months 2–3
Course pages + schema
Structured syllabus, FAQ, Course schema. Curriculum team reviews outcome accuracy.
-
Months 4–5
Learning paths & AI
3 pillar pages, comparison content, AI-referral enrollment tracking, GSC iteration.
Oura Atlas
Every step runs through a safe, audited loop.
Data from GSC & GA4 becomes a brief, execution runs through dry-run and backup, impact is verified — not wild production changes.
Relate?
Is this your story?
A good fit if…
- Course platform with 20+ courses and stagnant enrollment
- Prospects research recommendations on Google & AI before checkout
- Curriculum team ready to write answer-first outcomes & syllabus
Less of a fit if…
- Only 3–5 courses — need positioning, not a content system
- Unwilling to publish detailed syllabus publicly
- Need enrollment spike in 2 weeks with discounts only
“We have great material in the LMS. But public course pages are posters — AI cannot cite posters.”
“Prospects from AI research already know which skill they want — enrollment conversion beats discount traffic.”
Measured impact
Impact
6 → 12
structured course pages
4
skill queries in AI shortlist
+45%
AI referral in GA4
- Curriculum has a public page standard — aligned with LMS material quality.
- Marketing reduces discount dependency — measured self-research channel.
- Student success: prospects arrive with clear expectations — churn down.
Squad
The team behind execution
Managed service means real people operating Atlas — not software sold as a self-serve product.
Technical SEO Lead
Core Web Vitals, architecture, schema, and overall site health.
Content & Editorial
Answer-first content that Google reads and AI engines cite.
AI Visibility Specialist
OAVF execution, query mapping, and citation monitoring across generative engines.
Explore more
Related case studies
Healthcare services / SMB
My new website looks great — why is nobody finding us on Google?
New clinic, 3 weeks live — 12 GSC impressions, 0 patients from Google
Oura Starter Read →
Professional services / Local B2B
Our team is capable — why do competitors always own page one and we never get named by AI?
Services agency: capable on page 2 — prospects already have a ChatGPT shortlist
Oura Domination Read →
First step · free
Want results like this for your brand?
Start with a Free Mini Audit — we map your fastest path.
- Free
- No commitment
- Results in 48h
Win on Google. Get cited by AI.
Step 1 / 3
1. About you
Continue →
SEO
3 priorities
GEO
citation gap
AEO
schema check
FAQ — Case Studies
What was the main challenge in this EdTech / Online education case study?
EdTech platform with 40 live courses, 12,000 active students, enrollment growth slowing. Average course page: 200 words of marketing — no structured syllabus, learning outcomes, duration, or level comparison. When prospects asked ChatGPT "best data analyst course in Indonesia" or "learn Python for beginners, which platform", competitors appeared — this platform did not. Affiliates and discounts raised enrollment temporarily; organic flat. High churn on older courses because expectations did not match thin descriptions. This story shows how the Oura squad + Atlas handle similar profiles through the Oura Content package.
Which Oura package ran the "EdTech: 40 courses live — ChatGPT recommends competitors, not us" project?
This project ran on Oura Content. Service scope was tailored to EdTech / Online education — see the timeline and approach sections on this page for execution detail.
What is the main takeaway from this case study?
In EdTech, a large catalog without citable course pages means invisible in AI recommendation. Syllabus is SEO. Figures on this page come from real client data with publication permission.
How long until impact usually shows?
It depends on technical foundation and competition. Technical fixes and quick wins often show in 4–8 weeks; organic momentum and AI referral usually need several months of continuous iteration.
Which package fits a profile like this?
See the sidebar summary — each case study links to the Oura package that ran it. A Mini Audit helps map the realistic tier for your context.