How resolving upstream tracking vanity signals dropped acquisition cost from ₹160 to ₹36—while quadrupling genuine student activations.
Internshala is India's leading career acceleration and internship platform. When platform ad algorithms drifted into harvesting low-intent, zero-conversion vanity registrations, Abhishek rebuilt the entire upstream attribution telemetry and downstream activation loop.
-77.5% net acquisition spend
True ledger-verified acquisition cost per fully authenticated, enrolled candidate.
Verified course progress
Increase in students completing their first professional module within 7 days of signup.
Eliminated phantom spend
Budget redirected away from bot-heavy mobile interstitial placements and ghost traffic.
Audit reconciliation match
Full alignment between Meta Ads Manager reporting and raw production PostgreSQL logs.
The Illusion of Marketing Efficiency
In rapid-scale consumer education and career platforms, ad algorithms are hyper-adept at giving marketing leaders exactly what they configure—even if it destroys unit economics. Internshala’s ad campaigns were reporting massive numbers of registrations at record-low costs.
“When an ad algorithm finds out you value form submissions over actual classroom logins, it will find you millions of people who love tapping forms and will never study.”
— Abhishek Pal, Architectural Post-Mortem
The discrepancy surfaced inside weekly cohort reviews: customer acquisition costs appeared artificially low, yet paid course starts and active application submissions were completely stagnant. Marketing reported green dashboards; operations saw empty student seats.
Why The Algorithm Went Rogue
- closeShallow Trigger: Conversion pixel was fired on step-1 form input, counting accidental keystrokes and abandoned leads as victories.
- closeIn-App Interstitial Cannibalization: Machine bidding gravitated toward gaming app banner placements where mobile users tapped mistakenly.
- closeAudience Overlap: 34% of paid spend was inadvertently retargeting already-logged-in students searching for internship listings.
From discovery audit to completely clean ledger attribution
The 3-Phase Turnaround Sequence
Abhishek deployed a disciplined, step-by-step decoupling process that separated ad network vanity signals from core business operations.
Forensic Telemetry & Holdout Audit
Mapped every single conversion tag across Google Tag Manager, Meta Pixel, and GA4 against Internshala’s Postgres database. Instituted clean geographical holdout markets to establish pristine baseline organic demand without paid media contamination.
Upstream-to-Downstream Rewiring
Swapped platform bidding endpoints from browser page-submits to back-end Conversions API (CAPI). The ad algorithms were now forced to optimize for users who not only verified their accounts via OTP but completed at least 3 minutes of career training.
Exclusion Architecture & Clean Retargeting
Engineered dynamic real-time suppression lists. Existing registered users visiting for routine internship searches were instantly excised from paid ad auctions. Bidding capital was strictly concentrated on high-intent new prospective learners.
Explore the Forensic Inspection Ledger
Inspect the complete raw side-by-side data, production event pipeline schema, and audited unit economics table in the full-screen forensic modal.
Does Your Ad Engine Suffer From The Same Illusion?
Growth-stage venture founders frequently pay a hidden 40%–80% efficiency tax due to broken conversion feedback loops. Ask yourself these three questions:
Divergent Performance Trajectories
Are your performance ad dashboards reporting record ROAS or declining cost per lead while your monthly collected revenues or bank balances fail to keep pace?
Unreachable & Disoriented Leads
Does your sales or customer success team report that leads generated from paid campaigns rarely answer emails, bounce phone verification, or have no memory of requesting info?
Bidding On Your Own Organics
When an existing, happy customer searches for your brand name or tries to log in, are they clicking on expensive sponsored ads because dynamic negative audiences aren't in place?
Have a similar tracking or efficiency breakdown in your growth stack?
Abhishek conducts bespoke 2-week Telemetry & Unit Economics Audits for high-conviction founders spending upwards of ₹1M/month on performance media.