How Strata Learn made organic search their cheapest enrolment channel in 8 months
An EdTech Series A had spent two years dependent on paid performance for enrolments. We built a compounding SEO and content system targeting the jobs-to-be-done of working data professionals — non-brand organic became the #2 channel by enrolment volume and the cheapest by CAC.
The challenge
Strata Learn had raised a Series A on the back of a strong paid acquisition playbook on Meta and LinkedIn. Blended CAC was ₹4,200 per paid enrolment and the board had green-lit the next tranche of spend. But the founders saw a wall coming. CPMs on LinkedIn had risen 38% year-over-year. Meta creative fatigue was hitting every 11-14 days despite a full-time in-house designer. They needed a channel that could compound rather than linearly scale with spend.
The organic footprint was bare. 14 indexed pages, domain rating 18, fewer than 40 referring domains. Their blog was a wasteland of product announcements. There was no category page, no hub, no internal linking to speak of. Worse, their canonical course pages were blocked by JS-rendered content that Googlebot was not reliably executing — two of their best courses did not rank for their own brand name plus course name combinations.
The mandate: a 12-month compounding content engine that could get non-brand organic to at least 20% of total enrolments.
Our approach
Phase 1 — Technical foundation and keyword strategy (weeks 1-4)
We moved all course pages to server-rendered HTML, fixed internal linking structure, and built out XML sitemaps by content type. Keyword research was built around three layers: TOFU (how-to and definitional queries at KD 20-35), MOFU (comparison and 'best for' queries at KD 30-50), and BOFU (brand + course and pricing queries). We shipped with a 160-keyword target map spread across 12 topical clusters.
Phase 2 — Pillar and cluster content build (months 2-5)
Six pillar pages went live — one per core skill track (SQL for analysts, Python for data science, dbt for engineers, etc.). Each pillar had 8-12 supporting cluster articles linking up to it, and each cluster article linked laterally to 2-3 peer articles. Total: 82 articles, all authored by practitioners with named bylines and credentialed author pages for E-E-A-T.
Phase 3 — Comparison and tool content (months 3-7)
The conversion magic was in MOFU content. We built 18 comparison pieces ('Strata Learn vs Coursera for SQL', 'DataCamp alternatives for working professionals') and 11 free interactive tools (SQL sandbox, Python exercises with graded feedback). The tools attracted backlinks organically — 73 referring domains in six weeks — and the comparison pages converted at 4.8% page-to-enrolment.
Phase 4 — Measurement, iteration, and topical expansion (months 6-8)
We instrumented every article with scroll depth, read-rate, and exit-to-course tracking. Articles with low read-rate were either consolidated or rewritten. We used GSC's 11-20 position data to find articles one optimisation pass away from page one — 23 of those were revised with better intent match, better titles, and clearer H2 structure, and 17 moved into top-10 positions within the next crawl cycle.
The results
Month 8, attributed via last non-direct click in GA4.
Against blended paid CAC of ₹4,200.
Ahrefs tracked set, India + US.
Organic backlinks from tool content, no outreach.
GA4 organic channel, excludes brand.
Cleared the board's 20% threshold at month 7.
“Organic was an afterthought for us. Eight months of compounding content and technical SEO later, non-brand organic is our #2 channel by enrolments and the cheapest by a wide margin. It is the work that has kept paying after we stopped spending on it.”
What we got wrong — and what we'd do differently
The first three pillar pages we shipped were too long and too comprehensive. We had written 4,000-5,000 word ultimate guides because that is what the SEO playbooks suggested. They ranked, but the read-through was 18% and enrolment rate from those pages was weak. We rebuilt them as 1,800-word tight pieces focused on a single job-to-be-done, with the depth moved into linked cluster articles. Rankings held and conversion tripled. The lesson: comprehensiveness is not the same thing as usefulness, and Google in 2026 can tell the difference.
We also underestimated how long it would take to find authors who could write with genuine authority on data engineering topics. Our first three writers were generalist content marketers and the output was thin. By month three we had replaced the roster with two practicing data engineers who wrote part-time for us at 4x the per-word rate. That was the single highest-leverage decision in the programme.
Services used in this engagement
Free Audit — No Commitment
See exactly where your marketing is leaking revenue — for free.
We'll analyse your SEO, paid ads, and conversion funnel and send you a prioritised action plan within 48 hours — whether you hire us or not.
Get Your Free Audit