The Hope That Was Priced In — What Venture Capital Did, and Did Not Do, to Indian Education

Pritiraj BrahmaFounding Partner, Pritiraj & Partners · 2 October 2026

Indian edtech raised about $4.1 billion in 2021 and $321 million in 2023, on Tracxn’s count. Byju’s, once valued at a reported $22 billion, entered insolvency over ₹158.90 crore owed to the BCCI. PhysicsWallah listed in November 2025 and is putting IPO money into physical centres. Venture capital brought cheap, wide access and a sales machine pointed at children. India regulates it as a consumer product, not as education.

ProvenanceFunding figures are from Tracxn’s EdTech India reports (August 2024 and as reported in 2026); Tracxn has revised its annual series between editions, and both versions are stated. Company figures are from PhysicsWallah’s offer documents as reported at listing (November 2025) and from tribunal and court records on Byju’s (NCLT, July 2024; Supreme Court, October 2024). Regulatory references are the Ministry of Education guidelines of January 2024 and CCPA guidelines of November 2024.

In 2021 Indian education technology raised about $4.1 billion, on Tracxn’s count. By 2023 the figure was $321 million. Between those two numbers sits the largest experiment India has run in letting private risk capital into the business of teaching its children. It produced something real: learning at a price far below what a family used to pay a tutor, reaching towns the old coaching industry never served. It also produced a sales machine pointed at parents, a company once valued at a reported $22 billion in insolvency over a ₹158.90 crore sponsorship debt, and a regulatory answer that treats the whole sector as a consumer-protection problem. Nobody was asked to show whether the children learned more.

Venture capital has largely not bought Indian schools, and there is a structural reason. A school in India must, as a rule, be run by a not-for-profit trust, society or company, so the equity that a venture fund buys cannot easily sit inside one. Capital therefore went to the parts of education that sit around the school: learning apps, test preparation, tutoring, and, more recently, chains of physical coaching centres. The child still attends a regulated school in the morning. Increasingly, what decides their examination result happens in the unregulated hours after it.

Six findings anchor this analysis:

1. The money arrived in a single wave and left almost as fast. Tracxn’s August 2024 report records $4.1 billion raised by Indian edtech in 2021, $2.4 billion in 2022 and $321 million in 2023, an 87 per cent fall in a year. A single quarter, Q3 2021, brought $2.48 billion. Tracxn’s later series, as reported in 2026, puts 2021 at roughly $4.3 billion and 2023 at $265 million, and counts funded rounds falling from 368 in 2021 to 36 in January to August 2026. The editions differ in detail; both describe a boom and a collapse inside three years.
2. The capital concentrated on school-age children and examinations. On Tracxn’s figures K-12 edtech had raised about $8 billion to August 2024, and test preparation about $7 billion. Its 2026 series puts K-12 at 51 per cent of all sector funding between 2021 and 2026. Investors were not, for the most part, funding universities or adult skills. They were funding the competition for marks and seats.
3. The largest failure came through a creditor, not a regulator. On 16 July 2024 the National Company Law Tribunal admitted the BCCI’s insolvency application against Think and Learn, Byju’s parent, over unpaid dues of ₹158.90 crore. On 23 October 2024 the Supreme Court set aside the appellate tribunal’s approval of a settlement. No education authority had any formal role in the collapse of what had been the country’s most valuable education company.
4. The successful model is moving offline. PhysicsWallah listed on 18 November 2025 at an issue price of ₹109. Its offer documents, as reported, showed revenue rising from ₹744 crore in FY23 to ₹1,941 crore in FY24 and ₹2,887 crore in FY25, losses of ₹84 crore, ₹1,131 crore and ₹243 crore over the same years, 4.46 million paid users in FY25, and 198 offline centres at 31 March 2025. ₹460.55 crore of the fresh issue was earmarked for new offline and hybrid centres. Public-market money is now building physical coaching capacity.
5. The rules that exist are about advertising, not learning. The Ministry of Education’s guidelines of January 2024 bar coaching centres from enrolling students under 16, require tutors to be graduates, prohibit guaranteed ranks and misleading promises, and propose penalties of ₹25,000 and then ₹1 lakh before deregistration. Implementation was left to the states. The Central Consumer Protection Authority issued guidelines on misleading coaching advertisements on 13 November 2024, and by April 2025 had issued 49 notices and imposed ₹77.60 lakh in penalties on 24 coaching centres.
6. No public body measures whether any of this improves learning. The school system has national learning surveys. The parallel system that families now pay for has no published outcome measure at all: not completion rates, not score improvements, not refund rates. The only numbers in the public domain are funding rounds, valuations, user counts and penalties.

The hope

The hope was real and it should be stated plainly. Before venture-funded edtech, serious examination preparation in India meant moving to a coaching town or paying a premium in a metro. The new model put competent teaching on a phone at a fraction of that price. PhysicsWallah’s growth from a YouTube channel to 4.46 million paid users is, in part, a measure of how much unmet demand existed in towns the old industry never reached. For a family in a district headquarters, a low-cost course with a recognisable teacher is a genuine widening of opportunity.

The capital also forced the incumbents to change. Price, reach and the use of video at scale became normal in a sector that had been local and expensive. Whatever happens to individual companies, that shift is unlikely to reverse.

The scare

The scare is in what the money was paid to maximise. Venture capital rewards growth, and in education growth means enrolling more children, earlier, at higher lifetime value. The Ministry’s January 2024 guidelines exist because the government received complaints about student suicides, fire incidents, poor facilities and teaching practices in coaching centres. The under-16 bar is a direct response to the push to enrol children younger. The CCPA guidelines exist because guaranteed ranks and selective use of toppers’ photographs had become a standard selling tool.

The Byju’s collapse showed a second risk. When a venture-funded education company fails, the people exposed are not only its investors and lenders but families who paid in advance for courses, and employees owed wages. The insolvency process is designed to rank creditors, not to protect a child’s course of study. There is no education regulator with a duty to step in.

Why the regulator is a consumer authority

India’s education law regulates institutions: a school is recognised, a university is created by statute, a college is affiliated. A learning app is none of these. A coaching chain is a business. So the state reaches the sector through the instruments it has, which are consumer protection, advertising standards and, when a company fails, insolvency law. Each treats the student as a customer. None asks what was learned.

The shift offline makes this gap more consequential, not less. A physical coaching centre with hundreds of students in a building is exactly what the January 2024 guidelines were written for, and their force depends on each state adopting them. A national company now opens centres faster than many states have notified rules to govern them.

The counter-case

The strongest argument against regulating further is that the market has already disciplined the sector. The funding collapse punished the companies that grew without a business model. The survivors are leaner, cheaper and closer to what families want. PhysicsWallah’s losses fell from ₹1,131 crore to ₹243 crore in a year; a listed company publishes audited accounts and answers to a securities regulator. Adding an education regulator on top, the argument runs, would raise costs and price out exactly the low-income students the model reached.

Much of that is right, and nothing here argues for licensing teaching. But market discipline measures whether a company makes money, not whether a child learns. A securities filing discloses revenue and paid users; it does not disclose how many students finished a course or improved a score. The gap is not regulation of price or entry. It is the absence of any published outcome, which is the one thing a family cannot find out for itself before paying.

What we would do

1. Publish an outcome disclosure for every registered coaching provider above a size threshold. Completion rates, refund requests and refunds paid, and, where a provider advertises results, the denominator: how many students enrolled for every one whose result is advertised. The CCPA guidelines already restrict how results are shown; this would require them to be shown honestly.
2. Protect advance fees. Above a defined amount, fees paid in advance for a course should be held in a protected account or covered by insurance, so that a provider’s insolvency does not take a family’s money with it. Families are among the most exposed creditors in an education company collapse and currently have no special standing.
3. Finish adoption of the 2024 coaching guidelines in every state, with a public tracker. The Ministry of Education should publish which states have notified binding rules, which have not, and how many centres are registered in each. National chains open centres in every state; the floor should be the same in every state.
4. Bring large offline coaching centres under the school safety regime. Fire safety, building occupancy, counsellor access and the mental-health duties the Supreme Court set in July 2025 should apply to a coaching centre with hundreds of students exactly as they apply to a school of the same size.
5. Measure the shadow system. Add questions on paid coaching, by type and spend, to the national household and learning surveys, so that the state can see how much of a child’s education now happens outside the school and whether it is closing or widening gaps.

Venture capital did not create the demand for coaching. It found it, priced it, and briefly bet that it could grow without limit. The demand outlasted the bet. The task now is to make sure the next wave is judged on what reaches the child, not on what the funding round announced. Politics is the promise. Governance is the delivery. In education, delivery is the only number that should ever have been priced in.

Companion essay. The same question, approached from the other side: The Count the Campus Does Not Keep — Student Suicide and the Institutions That Do Not Report It.