A lot of JEE Prep instructors moving online are excellent teachers who have spent years in a physical coaching classroom, and that experience is exactly what makes some of their online mistakes surprising, because the things that worked in a classroom of sixty students do not automatically survive the move to a screen watched by someone at 11pm three states away. The mistakes below show up again and again in JEE Prep courses specifically, not because these instructors are careless, but because the offline playbook and the online one solve different problems, and a mistake that costs you nothing in a physical classroom can quietly cap your completion rate at 30% online without you ever seeing why.
Pricing it like a PDF, then picking a platform that quietly matches
The single most common mistake is anchoring price to what a printed notes package or a single test series used to cost, instead of what a full structured prep program with video, tests, doubt support and a certificate is actually worth. Students comparing a low priced course to a ₹45,000 offline institute fee will happily pay more than a token amount once they understand they are getting a structured two year program, not a PDF dump, and instructors who underprice out of habit end up attracting students who treat the course as disposable and never finish it. The framework in pricing your course at 999 vs 1999 vs 4999 is worth working through specifically for JEE, where the real competitor set is offline institutes charging five figures, not other online creators racing each other to the bottom. This underpricing habit tends to travel together with a second mistake, choosing a platform that takes no upfront fee and instead takes a cut of every sale, which feels free right up until a single high selling cohort launch and the instructor realizes a meaningful slice of a strong month went straight to the platform. Over a year of consistent JEE cohort launches, that commission usually adds up to several times what a flat annual platform fee would have cost, and the math is laid out in the real cost of free course platforms if you want to run your own numbers against your typical batch size, since the two mistakes together, an underpriced course on a commission-heavy platform, compound in exactly the direction that hurts most.
Recording a live class and calling it a course
A ninety minute live Zoom recording of a Physics lecture, unedited, with the first ten minutes spent on attendance and announcements, is not a course module, it is a recording of a course module, and students treat it accordingly by abandoning it halfway. The instructors whose completion rates hold up chop the live recording into twelve to eighteen minute concept segments, cut the dead air, and release them as a structured module rather than a raw feed, which takes real editing time but is the difference between content students watch at double speed and content they skip past entirely. It also changes how the content ages, since a well cut concept segment stays useful to next year's cohort with only minor updates, while a raw ninety minute recording tied to a specific batch's live questions and a specific date rarely gets reused at all, which means the editing time you skip this year quietly becomes work you have to redo from scratch next year.
Skipping the error log entirely
JEE prep lives and dies on a student's ability to see their own weak topics clearly, and a course that delivers the same linear sequence to every student regardless of their test performance is missing the single highest leverage tool available, a simple weak topic tracker built from test results. This does not need to be complicated engineering, even a downloadable worksheet that a student fills in after every test, listing which question types they missed and why, functions as a personalized path without any adaptive technology at all, and for JEE specifically it works best organized by error type, a silly mistake, a concept gap, or time pressure, rather than by chapter alone. Instructors who skip this step tend to assume a student who scores poorly on a mechanics test just needs more mechanics videos, when the actual gap is often time pressure carried over from a slow start on an earlier section, a distinction only visible once you are tracking error type rather than just chapter and score. Over a full season this log also becomes a genuinely useful teaching tool for you, since a pattern of the same error type showing up across dozens of students in the same cohort is usually a sign the concept itself was explained in a way that invites that mistake, which is far more actionable feedback than a generic end of module rating ever gives you.
- Price for the full 2-year program, not a single test series
- Edit live recordings into short concept modules before publishing
- Build a weak-topic error log into every test cycle
- Set a clear refund window before your first cohort launch
- Ask your strongest scorers for a referral instead of waiting for one
Refund policy written after the first refund request, not before
JEE Prep courses attract a particular kind of buyer, the parent purchasing on behalf of a dropper who might switch to an offline institute six weeks in, or a student who buys during a launch discount and then requests a refund after watching two videos. Writing your refund window, what counts as started for the course, and how partial cohorts are handled before you ever launch avoids the awkward improvisation that happens when the first request lands in your inbox, and the policy structures worth adapting are in refund policy for course creators. A policy decided calmly in advance, and stated plainly on your sales page rather than buried in fine print, also does quiet work on trust, since JEE parents specifically tend to read a clear refund window as a sign the instructor is confident in the program rather than hiding behind vague terms. Instructors who improvise their first refund decision under pressure, mid-launch, with a frustrated parent in their inbox, often end up setting an informal precedent that later requests then expect, which is a far worse outcome than a slightly stricter policy stated clearly from day one.
Relying on one discovery channel and no referral loop
Many JEE Prep instructors get their first hundred students from a single YouTube channel or Telegram group and then plateau there, without ever building a deliberate referral loop from their strongest scorers, even though a JEE topper mentioning your course by name is worth more to a prospective student than any ad. A simple structured ask, for instance a message to your top decile of test scorers each quarter asking them to share their result and tag your course, systematically outperforms occasional organic mentions, and the mechanics of building this loop are covered in turning course buyers into referrals. Diversifying beyond a single channel matters too, since JEE aspirants and their parents research heavily before buying and respond better to visible proof spread across several places than to reach concentrated in just one, and an instructor who has only ever been discoverable in one Telegram group finds out how fragile that plateau was the moment that one channel's engagement dips for unrelated reasons.
Every one of these mistakes is fixable without a platform migration or a production overhaul, and if you are building or rebuilding a program specifically for JEE aspirants, a course platform built for JEE Prep instructors removes at least the structural excuses, leaving you to focus on the sequencing, editing and error tracking work that actually moves completion rates.