What Happened to Chegg? How Free AI Broke Homework Help

Chegg lost most of its business because generative AI broke both halves of its model at once. Free chatbots such as ChatGPT gave students instant step-by-step answers, which removed the reason to pay for Chegg's library of solutions. Google's AI Overviews then answered questions on the search results page, which cut the free traffic that had brought students to Chegg in the first place. Revenue fell from about $776 million in 2021 to $376.9 million in 2025.

The turning point came on a single earnings call. On 1 May 2023 Chegg told investors that ChatGPT was hurting new sign-ups. The next day its shares fell about 48%. Two years earlier, in February 2021, the stock had traded above $110 and the company was worth about $12–14 billion, the most valuable US consumer edtech stock of the pandemic.

What was Chegg's business model?

Chegg began in 2007 as an online textbook rental company. Students paid a fraction of the list price to rent a book for a term and sent it back in a prepaid box. The model needed heavy capital, because Chegg had to buy books before it could rent them. Dan Rosensweig, hired as CEO in 2010, saw textbooks as the acquisition channel and digital services as the business.

The pivot took five years. Chegg bought Cramster, a homework-help site, in 2010, and built Chegg Study. In 2014–15 it moved its print inventory to Ingram Content Group, keeping the student relationship without owning the books. By 2019 services were more than 80% of revenue.

Chegg Study cost about $14.95–$19.95 a month for most of its history. It sold expert-written textbook solutions and an "ask an expert" Q&A service. Tens of millions of answers were written by contract experts, many in India, paid per answer. Once written, an answer could be sold to every later student with the same textbook problem.

The acquisition engine was free. A student typed a textbook problem into Google and landed on a Chegg page showing a blurred answer behind a paywall. Gross margins on services ran above 70% and adjusted EBITDA margins reached about 30–40% at the peak (reported). Customer acquisition cost was low because Google did the acquiring.

How did the pandemic change Chegg?

Remote learning lifted subscriptions sharply. Revenue rose from about $411 million in 2019 to about $644 million in 2020 and about $776 million in 2021. Subscribers peaked at about 7–8 million in 2021 (reported).

The growth came with a reputation problem. A 2021 study in the International Journal for Educational Integrity found a roughly 196% increase in STEM questions posted to Chegg during the early pandemic compared with the year before, many apparently from timed assessments. Forbes ran a cover story the same year with the headline "This $12 Billion Company Is Getting Rich Off Students Cheating Their Way Through Covid." Chegg responded with an honor code and cooperation with university investigations. Faculty largely kept their view.

Chegg also made its best financial decision of the decade in this period. It raised about $1 billion of 0% convertible notes in August 2020, on top of earlier low-coupon notes. That cash later kept it solvent.

What happened to Chegg after ChatGPT?

The decline ran in stages, each one larger than the last.

Subscribers fell to about 2.6 million by the second quarter of 2025, down about 40% from a year earlier, and Chegg stopped disclosing the figure afterward. Its 2025 filings attribute the 43% drop in Academic Services revenue mainly to "reduced traffic which led to fewer subscribers." Gross margin fell from 71% in 2024 to 60% in 2025, and the company guided to 48–49% for the third quarter of 2026, because the fixed costs of content and technology did not fall as fast as revenue.

CheggMate and the OpenAI partnership

Chegg saw the threat early. On the same May 2023 call it announced CheggMate, built with OpenAI's GPT-4. The disclosure was honest. The partnership, as the book notes, strengthened the supplier that also sold directly to students. A model provider that can produce a step-by-step answer for almost nothing does not need an intermediary to sell it.

The Google lawsuit and strategic review

From 2024 Google's AI Overviews began answering queries on the results page. For Chegg this was worse than a competitor: it was the channel owner choosing to answer the question itself. In February 2025 Chegg sued Google over AI Overviews and announced a strategic review, advised by Goldman Sachs (reported).

The review found no acceptable buyer. On 27 October 2025 Chegg said it would stay independent, cut about 45% of its workforce (388 roles), and bring back Rosensweig as CEO. Nathan Schultz, CEO since June 2024, stepped down to become an adviser.

Where does Chegg stand in late 2026?

Chegg is smaller, solvent and still shrinking. Full-year 2025 adjusted EBITDA was $68.5 million on $376.9 million of revenue, with a GAAP net loss of $103.4 million. In the second quarter of 2026, revenue was $51.8 million, adjusted EBITDA $9.1 million and the net loss about $3 million. Cash and investments were about $72 million at 30 June 2026, against about $33.8 million of 2026 notes due to be repaid in the third quarter.

The company is repositioning toward "skilling": Busuu language learning and Chegg Skills, sold to consumers and employers. That line was about $69 million in 2025. Skilling grew 11% in the fourth quarter of 2025, but growth slowed to 2% in the second quarter of 2026, short of the double-digit target set in October 2025. Busuu, bought at the 2021 peak and partly written down, became one of the few lines not directly damaged by AI answers.

Why couldn't Chegg defend itself?

Chegg's moat had two parts: a proprietary library of answers and a strong search ranking. Both depended on scarcity that no longer existed.

The library had been expensive to build and cheap to serve. A general model made it cheap to build too. Rivals such as Course Hero, Brainly and Quizlet faced the same pressure, but Chegg was the largest and most dependent on paid answers.

The search channel had always belonged to someone else. The book's verdict is that "Chegg had content but not the default channel to students; Google and OpenAI had both." Chapter 10 of the book draws the wider line: tutoring promises learning, test prep promises a score, and homework help promises a finished assignment. A chatbot delivers a finished assignment instantly and free, so homework help was hit first and hardest.

The research also notes a decision that cost Chegg more than it needed to. It cut staff in rounds in 2023, 2024 and May 2025, each overtaken by further decline, before the 45% cut in October 2025. One deep cut early would have been cheaper.

Which factor of the learning equation broke?

The learning equation multiplies four factors: learning design, a willing payer, an affordable route to that payer, and evidence. Chegg's collapse turned on the second and third.

The payer factor went toward zero when students found a free substitute. They had paid for a finished answer, not for learning, and a chatbot gave them the finished answer. The route factor broke when Google answered queries itself. The learning-design factor had always been weak: worked examples are well supported by cognitive science when studied before solving similar problems, but Chegg never published independent controlled studies of Chegg Study, and many students used it to skip the work. A product sold for answers rather than understanding had no defense when answers became free.

This is the opposite of how Duolingo makes money. Duolingo also sells a consumer subscription, but it grows mostly through its own brand and habit loops rather than someone else's search engine.

What are the lessons from Chegg?

• A content library is not a moat when a general model can produce the content free. Ask what the next model release will give away.
• Own the channel, or know who does. Building acquisition on free search traffic left Chegg exposed to a single company's product decision.
• Raise cheap capital at the peak. Chegg's near-zero-coupon convertibles were its best financial move and the reason it avoided a debt restructuring.
• Cut once, and deeply, when the decline is structural.

Chegg sits alongside 2U and Byju's on our list of the biggest edtech failures, and its story frames the questions in our explainer on the AI tutoring business model. For the broader picture of how consumer and institutional edtech make money, see our edtech business model guide.

The takeaway

Chegg was a well-run subscription business for a decade, built on expensive-to-make answers and free search traffic. Generative AI made the answers free and Google's AI Overviews took the traffic, so both assets lost their value at the same time. Cheap capital raised at the peak kept Chegg solvent, but its academic revenue was still falling by about half a year in 2026. Chapter 10 of The Learning Equation explains why homework help was the first edtech category broken by AI.