A fast-growing online upskilling platform with high-AOV courses and a telesales team closing on calls. Meta and Google had never seen a single enrollment. They kept optimizing for the cheapest leads — which were also the least likely to become students.
The story in 10 seconds
The ad platforms had never seen a single enrollment, so they kept buying the cheapest leads — and 7 of 10 were unreachable or had no intent. We fed CRM enrollment data back to Meta & Google, split campaigns per course, and scored leads by payment intent. In 8 weeks, counsellor connect rate climbed from 38% to 61% and the cost per enrolled student dropped 29% — same budget, same in-house team.
01 — The Setup
High-AOV upskilling courses with a telesales team converting leads to enrollments. Meta and Google were tasked with volume — and they delivered. Lead numbers in the dashboard looked healthy.
In the call centre: approximately 70% of leads were either unreachable, had zero intent, or had applied for the wrong program entirely. Every counsellor hour spent dialing these was an hour not spent on genuinely interested prospects.
The platform’s job ends at the form fill. The real conversion work starts after it — and ads had no signal from that world at all.
02 — What Was Broken
What 100 leads become at each stage
Enrollment CAC declining month on month ($)
03 — What Flable AI Did
CRM enrollment records piped back into Meta and Google in real time. Platforms finally had signal on what a paying student looks like — not just a form-fill.
Optimization events shifted: Lead → Connected → Demo Attended → Enrolled, with dollar enrollment value passed back. Platforms stopped rewarding form-fillers.
Each course — Data Science, DevOps, Product Management and others — got its own campaign with separate audience logic, bids, creatives, and optimization signal.
Payment-intent signals scored and pushed back to ad platforms continuously. High-intent leads re-targeted; low-intent leads filtered before ever reaching a counsellor's queue.
04 — Outcome by Month 2
Jump in pipeline quality. Counsellors now spending the majority of their time on prospects who enrolled.
From 38% to 61% — nearly 2x more conversations with people who actually picked up.
Cost per enrolled student — the only metric that matters in an EdTech lead-gen model.
1 winner scaled 3x. 2 underperformers paused. First time program-level ROI was measurable.
Connect rate, lead-to-SQL, and enrollment CAC all trending correctly
Bottom line
When ad platforms can see who actually enrolls, they stop sending you people who never will.
Frequently Asked Questions
Because CPL prices a form fill, not a student. With ~70% of leads unreachable or no-intent, a cheaper lead just means cheaper junk filling the counsellor queue. In a lead-gen-to-telesales model the number that matters is cost per enrolled student — and that was climbing even while CPL looked great.
Once enrollments and their dollar value flow back to Meta and Google, the platforms optimise toward people who actually become students instead of people who fill forms. The algorithms finally learn from the outcome that matters, not the cheapest signal available.
No — it makes their spend smarter. Flable closes the loop between your CRM and the ad platforms and splits campaigns per program, so your team can finally see which courses acquire students profitably. The team keeps running growth; they just run it on enrollment signal instead of form-fill cost.
Low-intent leads get filtered before they reach the dialing queue and high-intent leads get prioritised. Connect rate rose from 38% to 61%, so counsellors spend their day on people who pick up and enroll — not chasing dead numbers.
A connection to your ad accounts (Meta, Google) and your CRM enrollment / payment data. From there Flable reconciles the full funnel — lead → connected → demo → enrolled — automatically, with no manual spreadsheet stitching.
Free 30-minute audit. We connect your CRM enrollment data to your ad platforms and show you what your actual cost per enrolled student looks like — on your own numbers, with your in-house team in the room.
Book a free audit →Setup in 48 hours · No credit card required · Backed by Microsoft & NVIDIA
Premium Electronics Retail
A premium electronics retailer hit its lowest-ever CPL while 70% of leads never visited a store. Connecting in-store data and flipping the signal to store visits and purchases cut acquisition cost 31%.
Luxury Fragrance D2C
A luxury fragrance D2C brand paid its agency 15% commission on inflated, Meta-attributed sales. An independent attribution layer exposed the real numbers — recovering ~$3K/month and lifting POAS 41% in two months.