The numbers tell the story faster than the press release
Chegg reported second quarter 2026 revenue of $51.8 million, down 50.7 percent from $105 million in the same quarter a year earlier. Adjusted EBITDA came in at $9.1 million, a 17.4 percent margin the company reached mostly by cutting costs rather than growing revenue. Non-GAAP operating expenses nearly halved year over year to $32.3 million. Guidance for the third quarter calls for another sequential drop, to $43 to $44 million in revenue and just $1 to $2 million in adjusted EBITDA. The stock closed near $0.88 after the report, down almost 15 percent on the day.
None of this is a one-quarter stumble. Chegg has spent over $14 million on severance in the first half of 2026 alone, part of a restructuring the company says will cut annualized costs by $100 to $110 million. Cash and investments stand at $72.3 million, with a net cash position of $38.5 million. That is a company managing decline carefully, not one about to run out of runway tomorrow. But it is also a company whose core product, on demand homework help and textbook Q&A, has been effectively replicated for free by ChatGPT, Gemini, and every other consumer chatbot a student already has open in another tab.
Why this vendor specifically got run over
Chegg's moat was never proprietary content in the way a database or a compliance engine is proprietary. It was aggregation and convenience: more than 100 million question and answer pairs, indexed and searchable, that a student would otherwise have paid a tutor to produce. Generative AI does not need Chegg's dataset to replicate that value. A model trained on the open internet already answers the same class of question, instantly, without a subscription. Former CEO Nathan Schultz described Chegg's prior identity as a transformation 'from a textbook rental business into a global learning platform driven by AI.' The irony is that AI became the product's replacement rather than its upgrade path.
This is a structural failure, not a marketing failure. Chegg tried content partnerships, price cuts, and product bundling, and none of it mattered because the substitute was free, fast, and already distributed to hundreds of millions of devices. The lesson generalizes well beyond one struggling edtech company. Any vendor whose value proposition is 'we answer a well defined question faster than you can search for it yourself' is exposed to the same dynamic, whether it sells to students, call center agents, or junior analysts. The tell is whether the vendor's product is a wrapper around information retrieval or a wrapper around a genuinely hard workflow.
The pivot, and why it deserves a fair hearing
Chegg's answer is to stop competing on content and start competing on outcomes. Rosenzweig is steering the company toward what it calls the employability market, using its dataset and AI tooling to help students find internships and jobs rather than just answer homework questions. Early signals include more than 10,000 beta testers for a new job matching platform, a soft launch planned for the third quarter of 2026, and six new partnerships under the Chegg Skills brand this year. The board's decision last October to remain an independent public company, rather than sell or wind down, was itself a bet that this pivot has enough runway to work.
Whether it succeeds is a genuinely open question, and that is exactly the point for anyone evaluating a similarly exposed vendor. A pivot toward outcomes, workflow integration, or proprietary data that a general model cannot easily replicate is a credible response to AI commoditization. A pivot that just adds a chatbot skin to the same underlying product is not. The difference between those two moves is usually visible in the vendor's own roadmap language within two or three quarters, well before the financials confirm it one way or the other.
What this means for your next vendor renewal
If your organization licenses learning content, tutoring platforms, knowledge bases, or any tool whose core function is answering a defined category of question, Chegg's income statement is the risk model you should be running against that vendor today. Ask directly: what percentage of this vendor's value could a well prompted general purpose model replace for an employee with access to ChatGPT Enterprise or Copilot. If the honest answer is most of it, price that risk into the contract term, not just the renewal date. A 51 percent revenue drop in a year moves fast enough to strand a multi year commitment long before the next scheduled renewal cycle gives procurement a chance to catch it.
This also argues for shorter contract terms and tighter usage based pricing on any content or knowledge vendor through 2026 and 2027, even at some premium to a locked in annual rate. Enterprise buyers spent the last decade optimizing procurement for vendor lock in discounts. That logic assumed the vendor's product durability, not just its balance sheet, was stable. Chegg shows that assumption can fail inside four quarters for a company that was, as recently as 2021, valued north of $10 billion.
The talent and L&D angle CIOs should not skip
There is a second read of this story that matters specifically for enterprise learning and development leaders. If a well resourced, purpose built company like Chegg could not out-execute a free general purpose chatbot on tutoring and Q&A, internal L&D teams building similar homegrown tools face the identical exposure. Custom internal knowledge bases, onboarding FAQ bots, and compliance quiz tools built on the same 'answer a defined question' logic are candidates for the same disruption, just without a public earnings call to announce it.
The practical response is to audit which of those tools are pure information retrieval and which are doing something a general model genuinely cannot: enforcing a regulated process, tracking individual competency over time, or integrating with systems of record. The retrieval layer should probably be replaced with, or supplemented by, the same AI tools eating Chegg's business. The workflow and compliance layer is where L&D and knowledge management budgets should concentrate, because that is the layer with a defensible reason to still exist in 2027.
