E-learning is training delivered through digital technology, remotely or in blended form, on platforms that manage courses, learners and assessments.
It is one part of edtech. Politecnico di Milano’s EdTech Observatory defines edtech as the set of solutions, enabled by hardware and software, that support the learning process.
The sector is consolidating. On May 11, 2026, Coursera completed its combination with Udemy. At the same time, generative AI is eroding the business of companies that sold homework help: Chegg’s own filings say students increasingly see tools like ChatGPT as an alternative to specialised education products. This guide covers where e-learning came from, why nobody agrees on its size, who leads it and what AI changes for learners.
How did e-learning start?
PLATO, developed at the University of Illinois from 1960 on the ILLIAC I computer, was the first generalised computer-assisted instruction system. It also pioneered tools we now take for granted: forums, message boards, online testing, email and chat rooms.
The next step was the learning management system (LMS), the platform that organises courses, users and assessments. Docebo, one of today’s listed LMS vendors, was founded in 2005 in Macherio, Italy.
In 2012 massive open online courses, or MOOCs, captured global attention with the promise of broadcasting top professors’ lectures to anyone with a connection.
The pandemic turned e-learning from an option into a necessity. By mid-April 2020, schools were closed in more than 190 countries, affecting 1.57 billion learners: over 90% of students worldwide.
| Year | Milestone | Why it matters |
|---|---|---|
| 1960 | PLATO launches at the University of Illinois | First generalised computer-assisted instruction system |
| 2005 | Docebo is founded in Italy | Corporate learning platforms become a product category |
| 2012 | MOOCs go mainstream | Free university courses for anyone |
| 2019 | MIT study of 12.67 million enrolments | Only 3.13% complete (2017-18) |
| 2020 | Pandemic school closures | 1.57 billion learners at home |
| 2023 | Chegg shares collapse | First listed company openly hit by ChatGPT |
| 2026 | Coursera and Udemy combine | The sector consolidates |
Why didn’t MOOCs deliver on their promise?
Justin Reich and José Ruipérez-Valiente of MIT analysed every course MIT and Harvard ran on edX from 2012 to 2018: 5.63 million learners and 12.67 million course registrations. Their paper, published in Science on January 11, 2019, found three patterns:
- most learners never came back after their first year;
- growth was concentrated almost entirely in the world’s most affluent countries;
- low completion rates had not improved in six years.
In 2017-18, only 3.13% of all participants completed their courses. Reich’s explanation was that what people are willing to pay for is the credential, not the teaching. Providers followed the money: the field shifted towards helping universities run online master’s degrees for professionals.
That lesson still shapes the industry. Platforms that sell verified skills and credentials have held up better than those that sell content alone.
How big is the e-learning market?
There is no single reliable figure, and the reason is how the market is defined.
Two global estimates show the gap:
- About $142 billion. A forecast cited by Politecnico di Milano put the 2023 revenue of companies selling hardware and software for learning at this level.
- Over $400 billion. Italy’s Proxima Observatory, in its 2026 EdTech Report, puts total spending on digital tools in education above this figure.
One number counts what technology vendors earn; the other counts everything the sector spends. Treating them as the same thing is the most common mistake in coverage of this market.
Investment data is more telling. Global venture funding in edtech fell from $20.8 billion in 2021 to $10.6 billion in 2022, according to figures cited by the Milan observatory. Proxima reports that global investment in edtech startups rose from $2.4 billion in 2024 to $2.6 billion in 2025. The sources differ, but the signal is the same: the 2020-2021 bubble has deflated.
Demand is driven by employers, not schools. Corporate training accounts for roughly 65% of global spending on digital learning, according to Proxima.
| Source | What it measures | Value | Reference |
|---|---|---|---|
| Forecast cited by Politecnico di Milano | Global revenue of learning technology vendors | $142 billion | 2023 |
| Proxima Observatory (Enzima12) | Global spending on digital tools in education | Over $400 billion | ⚠️ year to verify |
| Figures cited by Politecnico di Milano | Global edtech venture funding | $20.8bn → $10.6bn | 2021 → 2022 |
| Proxima Observatory | Global investment in edtech startups | $2.4bn → $2.6bn | 2024 → 2025 |
Who are the biggest e-learning companies?
The market splits into four models: corporate platforms, course marketplaces, consumer apps and study help.
- Coursera and Udemy. The combined company reaches over 290 million learners, 18,000 enterprise customers and 95,000 instructors. According to Coursera’s quarterly filing, the consideration paid to Udemy shareholders at closing had a fair value of $673.6 million, with Coursera stock at $5.66.
- Duolingo. Revenue reached $748.0 million in 2024, up 41%. In the third quarter of 2025 it had 50.5 million daily active users. The company describes its app as feeling more like a mobile game than an education product.
- Docebo. Its second-quarter 2026 revenue was $68.7 million, up 13% year on year.
- Chegg. The opposite case. Its shares fell nearly 50% on May 2, 2023, after the company said students were using ChatGPT instead of paying for its study tools.
| Company | Model | Latest figure | Source |
|---|---|---|---|
| Coursera + Udemy | Course marketplace and credentials | 290M+ learners, 18,000 enterprise customers | Coursera, May 11, 2026 |
| Duolingo | Consumer language app | 50.5M daily active users (Q3 2025) | Shareholder letter |
| Docebo | Corporate learning platform | Q2 2026 revenue: $68.7M (+13%) | Earnings release, Aug 7, 2026 |
| Chegg | Study help for students | Q4 2024 revenue: $143.5M (−24%) | CNBC, Feb 2025 |
Is AI hurting e-learning companies?
It depends on what they sell.
Companies that sell answers are losing. In February 2025 Chegg sued Google, arguing that AI-generated summaries in search results were cutting its traffic and revenue. Its fourth-quarter 2024 revenue had already dropped 24%.
Companies that sell skills are growing. Coursera says that in 2026 someone enrols in a generative AI course every three seconds on average. Duolingo said in 2025 that its AI investments had helped it produce content faster.
The pattern so far: when AI makes an answer free, the value moves to learning how to judge that answer and to proving the skill.
How did Docebo go from Italy to Nasdaq?
Docebo was founded in 2005 by Claudio Erba. Its growth has clear milestones:
- 2012: first funding round, €2.4 million from Principia SGR;
- 2013: first US office, in Athens, Georgia;
- October 8, 2019: listing on the Toronto Stock Exchange, raising C$75 million;
- December 3, 2020: first day of trading on Nasdaq. The offering closed on December 7 with $165.6 million in gross proceeds, including the underwriters’ over-allotment.
The company’s journey is the subject of an INT News Interconnection talk with Alessio Frigeri. He joined Docebo in 2009, when it had six people, and led Global Operations for about 14 years, through the dual listing and past $100 million in annual recurring revenue.
In the second quarter of 2026, annual recurring revenue reached $255.1 million, up 9.5%. Its largest OEM customer, a company that resells the platform inside its own offering, fell from 8.4% to 2.5% of recurring revenue: Docebo now depends far less on a single client. Full-year 2026 revenue guidance is $274.5-276.5 million.
What about Italy?
Skills. According to Istat, 54.3% of Italians aged 16 to 74 had at least basic digital skills in 2025, up from 45.7% in the previous survey. The EU average is 60%. Italy posted the largest increase in the EU after Denmark. The education gap persists: in the 25-54 age group, 86.1% of graduates have at least basic skills, against 33.5% of those with lower qualifications. Online courses and learning are among the fastest-growing internet activities.
Market. Politecnico di Milano valued the Italian edtech market at about €2.8 billion in 2022. Proxima’s 2026 report puts it at €2.7 billion, with around 11,000 employees. The two figures use different scopes and years, so they cannot show a trend.
Startups. Proxima counted 388 Italian edtech startups in 2024-2025. In 2024 they raised €74 million, up 174% on 2023.
Do we still need to study in the age of AI?
The data says yes: more than before, but different things.
What employers say. The World Economic Forum’s Future of Jobs Report 2025, published on January 7, 2025, estimates that 39% of today’s skills will be transformed or outdated by 2030. For every 100 workers, 59 will need training by 2030. Analytical thinking remains the most sought-after core skill: seven in ten employers consider it essential.
What studies say. The evidence points in two directions.
- Harvard. A study of 194 students in an introductory physics course, published in Scientific Reports in June 2025, tested an AI tutor. Students who used it learned more than twice as much as in an active-learning class, and in less time. The tutor, however, was designed by instructors, had the correct solutions and was built to make students reason.
- MIT. A preliminary study from the MIT Media Lab split 54 participants into three groups to write essays. Those using ChatGPT showed the weakest brain connectivity. Only 18 people completed the final session, and the paper was not yet peer-reviewed.
Both studies are small and context-specific, and neither settles the question. Together they suggest a practical rule: AI helps learning when it forces you to reason, and takes something away when it reasons for you. Studying today means memorising fewer answers and getting better at judging them.
FAQ
What is the difference between e-learning and edtech?
Edtech covers all technology that supports education, from schools to companies. E-learning is the part that delivers training online through platforms and digital courses.
How big is the e-learning market?
Estimates range from about $142 billion, counting only technology vendors’ revenue, to over $400 billion, counting total spending on digital tools in education. The difference comes from definitions, not errors.
Who invented e-learning?
The first generalised computer-assisted instruction system was PLATO, developed from 1960 at the University of Illinois.
Is Docebo an Italian company?
It was founded in Macherio, Italy, in 2005. It is now headquartered in Toronto and listed on the Toronto Stock Exchange and Nasdaq.
Do we still need to study now that we have AI?
Yes. According to the World Economic Forum, 59 out of 100 workers will need training by 2030. What changes is the focus: less memorising, more judging answers.
Sources
- Docebo, Q2 2026 earnings release, August 7, 2026 (SEC Form 6-K)
- Docebo, releases of December 2 and 7, 2020, on its US offering (SEC Form 6-K)
- Docebo, company history page
- Coursera, press release of May 11, 2026, and Form 10-Q for the quarter ended June 30, 2026
- Axios, May 11, 2026
- Duolingo, shareholder letters for Q4 2024, Q1 2025 and Q3 2025, and Form 10-K for 2025 (SEC)
- Chegg, Form 10-K for 2025; Fortune, May 2, 2023; CNBC, February 2025
- Reich and Ruipérez-Valiente, “The MOOC Pivot”, Science, January 11, 2019; Inside Higher Ed; Times Higher Education
- Politecnico di Milano, EdTech Observatory press releases
- MilanoFinanza, March 25, 2026; EduNews24, August 2026 (Proxima 2026 EdTech Report)
- Istat, Cittadini e ICT 2025, via Agenda Digitale
- World Economic Forum, Future of Jobs Report 2025, January 7, 2025
- Kestin et al., Scientific Reports, June 2025; Harvard Gazette, September 2024
- Kosmyna et al., MIT Media Lab, 2025 (preliminary study)
- UNESCO Courier, July-September 2020
- PLATO system record, via EPFL Graph
- Wikipedia (Docebo corporate history), cross-checked against company filings
