By: Prempreet Singh, Founder & CEO, SkillBytes
For about a decade, the story of Indian edtech was a story of money. Huge funding rounds, huge marketing budgets, huge promises. Somewhere in all of that, a simple fact got lost. Most students never finished anything they started.
The numbers were brutal and widely known across the sector. Completion rates often sat below 15 percent. That means for every hundred students who signed up, downloaded, and logged in, more than eighty-five quietly drifted away. The industry’s response was usually to spend more, on ads, on discounts, on celebrity faces, to pour new students into the top of a leaky bucket.
It’s worth asking why the bucket leaked in the first place. The easy explanations were that students lacked motivation, or that the content wasn’t good enough. Both were wrong. Plenty of students were motivated. Plenty of the content was excellent, built by serious educators. The problem was never content or motivation. It was format and friction.
Think about what the old model actually asked of a student in a Tier 2 or Tier 3 town. Find the app. Download it on a phone that may already be short on storage, over a connection that isn’t always fast. Create an account. Verify it. Remember a password. Open the app the next day, and the day after that, building a brand-new habit from scratch. Every one of those steps was a small wall. For a student with a good device and reliable broadband, the walls were low. For everyone else, they stacked up fast. By the time learning could actually begin, a large share of students had already fallen off.
This is the part the boom-era model got backwards. It treated learning as an event you travel to, a destination with a front door, a lobby, and a sign-in sheet. But a student’s day isn’t built around destinations. It’s built around small gaps, the ten minutes on the bus, the wait before dinner, the stretch after homework when the brain is still half-on. Filling those gaps is a habit problem, not a content problem.
The size of that gap is striking. By some estimates a student is exposed to as much as a thousand hours of learning context in a year, but fewer than two hundred of those hours are actually guided. The rest leaks away, and a good chunk of it leaked into tools that asked too much before they gave anything back.
So what works now? The honest answer is unglamorous. What works is removing the walls.
The most useful shift in the last couple of years has been meeting students inside the apps they already live in, rather than dragging them somewhere new. Indian students spend hours a day inside WhatsApp. It’s already on the phone. It already works on a basic Android handset on a patchy connection. There’s no download, no new password, no fresh habit to build. The messaging window they’re already staring at becomes the classroom.
That single change quietly fixes most of what the old funnel got wrong. When a student can go from a link or a QR code to their first lesson in under thirty seconds, the drop-off that used to happen during setup simply doesn’t happen. There’s nothing to abandon before you begin.
This logic isn’t new to the people who’ve built for India’s mass market before. “Adoption at scale never comes from better content or better technology. It comes from removing friction,” says Prempreet Singh, Founder and CEO of SkillBytes, who spent fourteen years building consumer products for the next billion users across India’s payments shifts before turning to education. “The moment a product meets users where they already are, it stops feeling like a product and starts feeling like a habit.”
The second thing that works is matching the lesson to the gap. Long video lectures assume a student who has carved out an hour and won’t be interrupted. Most students don’t have that hour, or the quiet. Compressing a concept into a five-minute unit that feels complete inside a chat thread fits the way a phone is actually used, in bursts, between other things. You’re working with the student’s attention span instead of fighting it.
This is harder to build than it sounds, which is probably why the well-funded players skipped it. Writing a lesson that genuinely lands in five minutes, inside a message thread, takes more pedagogical discipline than recording a forty-minute video. It means co-creating tightly with experienced educators and keeping every unit board-aligned, so nothing is sacrificed for the sake of brevity. Short is not the same as shallow, and getting that distinction right is most of the work.
There’s early evidence the approach travels. SkillBytes, which built its model around exactly this idea of board-aligned content delivered as five-minute units inside WhatsApp, reports reaching over 520,000 active students within eight months of launch, and generating around 1.7 crore interactions, without the heavy marketing spend that defined the previous era. The more telling figure is engagement density. Students average around 32 messages per session, and interaction volume is growing faster than the user base itself. People aren’t just signing up. They’re coming back and doing the work, which is the behaviour the old model could never reliably produce.
None of this means the boom-era platforms were building junk. Much of the content from that period was genuinely strong. The failure was a delivery failure. The industry built a great library and then put it behind a door that a large part of the country couldn’t easily open. The small-town student wasn’t left out because the material was too hard or because they didn’t care. They were left out because the format demanded a fast phone, a steady connection, and the patience to climb a setup process before a single thing was learned.
The lesson from the boom, then, is almost embarrassingly simple. You don’t win by shouting louder or spending more at the top of the funnel. You win by removing the reasons people fall out of the bottom. Put the lesson where the student already is, shape it to the time they actually have, and keep the quality honest. Get the format and the friction right, and motivation and content, the things that were never really the problem, finally get their chance to matter.
