TheAudioLearning sells a CPC medical coding certification with placement assistance. Good product, real instructor, genuine placement record. What it did not have was a way to consistently turn strangers into enrolled students.
Over 120 days we built one. Here is what happened, with the numbers.
The headline result
| Metric | Value |
|---|---|
| Ad spend | ₹18.6L |
| Impressions | 12.4M |
| Leads | 15,000+ |
| Cost per lead | ₹126 |
| Qualified leads | 5,800+ |
| Admissions | 780+ |
| Revenue | ₹1.02Cr+ |
| ROAS | 5.5X |
Bars are drawn to relative scale, not to raw proportion — at true scale the last stage would be invisible, which is itself the point.
Why we did not sell the course directly
A certification course is a considered purchase. The buyer is often changing careers, frequently nervous about whether they can do the work, and comparing several institutes on price and placement claims.
Asking that person to buy from an ad is asking for a decision they are not ready to make. Cold-traffic direct-sale campaigns in this category burn money.
So the entry point became a free live webinar with Aparajita Sudarshan, the instructor — 18 years of experience, published author in the field. The offer was not "buy our course." It was "spend 90 minutes with the person who would teach you, for free."
That reframes the decision entirely. The prospect is not evaluating a purchase; they are evaluating whether this person is credible. Most of them concluded that she was.
The funnel, stage by stage
Bars are drawn to relative scale, not to raw proportion — at true scale the last stage would be invisible, which is itself the point.
Ad → registration. Creative led with the career-change anxiety directly: Career restart? It's never too late. No age barrier, joining certificate, placement guarantee. Registration through a native lead form, not a landing page — fewer steps, better mobile completion.
Registration → WhatsApp confirmation. An automated message fired immediately: seat confirmed, date and time, instructor name and credentials, joining link to follow, please confirm availability. This single automation did more for attendance than any creative change.
Qualification inside the form. Four questions sorted the list before anyone spent sales time: did you attend the webinar, preferred language, science background or not, and how soon do you want to enrol — right away, this week, this month, or just exploring.
That last question is the sales team's entire prioritisation system.
Webinar → enrolment. The session taught genuinely useful material and closed with the offer. Direct conversion campaigns then retargeted attendees who had not yet enrolled — warm audiences only, never cold.
What the campaign-level numbers showed
Campaign-level cost per lead ranged from ₹11.40 to ₹41.16 across webinar batches. Averages hide the range, and the range is where optimisation lives.
Individual webinar batches performed very differently:
| Campaign | Leads | Cost per lead |
|---|---|---|
| Webinar — 12 Jan | 1,062 | ₹11.40 |
| Webinar — 20 Dec | 263 | ₹41.16 |
| Webinar — 22 Feb | 331 | ₹31.44 |
| March batch | 535 | ₹16.56 |
The spread between ₹11.40 and ₹41.16 is not randomness. It is timing, creative freshness, and audience saturation. The 12 January batch hit an audience that had not seen the offer before, with a creative that was working. By late February, frequency had risen and costs followed.
The lesson we took: batch-based offers need creative rotation on the same cycle as the batches, not slower.
Audience insight that changed targeting
Age and gender breakdown showed 18–24 and 25–34 dominating, with women at 49% of results and a cost per result of ₹10.95 against ₹11.50 for men.
Nearly identical cost, but the qualitative signal mattered more: a large share of leads listed themselves as doctors, nurses, and allied health professionals looking to move into coding. That is a very different message than "start a career in healthcare" — these people were already in healthcare and wanted out of clinical hours.
Creative shifted accordingly, and CTR held at 4.1% across the full period.
What actually produced the revenue
Not the targeting. Not the creative, exactly. Three things:
- A free, high-credibility entry point that matched how people actually buy education.
- Automated WhatsApp follow-up that caught interest at its peak instead of the next morning.
- Qualification questions that let a small sales team spend its hours on the 5,800 who mattered rather than all 15,000.
The ads created the opportunity. The system converted it.
What we would do differently
Rotate creative faster against batch cadence. Build a nurture sequence for the "this month" and "just exploring" segments rather than letting them cool. And instrument the offline conversion loop earlier so the platform learned which leads actually enrolled, not just which ones filled the form.
The full campaign breakdown, with the dashboard screenshots, is in the TheAudioLearning case study.