Coursera 2026 Q3 Earnings: Revenue $376.7 Million, Project Helix Planned to Open in the First Half of 2027

On October 28, the US online education company Coursera (NYSE: COUR) released its financial results for the third quarter of 2026. This was the company's first full quarter after completing its merger with Udemy; both revenue and adjusted EBITDA beat market expectations, but next-quarter guidance fell short of expectations, causing the stock price to fall in after-hours trading.

Coursera's third-quarter 2026 revenue was $376.7 million, up 94% year over year and 2.2% above market expectations. The sharp revenue growth came mainly from Udemy's results, consolidated starting May 11, 2026; excluding the impact of the merger, revenue on an organic basis actually declined by about 3% year over year.

GAAP net loss was $25.6 million, significantly wider than the $8.6 million loss in the same period last year; GAAP loss per share was $0.10, compared with $0.05 in the same period last year. Adjusted earnings per share were $0.21, 16.9% above market expectations.

After the earnings release, Coursera's stock fell about 8% in after-hours trading to $4.27. The core market concern is that although the merger brought significant revenue growth and margin improvement, organic growth momentum (especially in the consumer business) remains weak, and short-term guidance failed to show sufficient signs of recovery.

Coursera CEO Greg Hart emphasized on the earnings call that the integration is proceeding faster than originally planned, and the company expects to achieve more than $150 million in annualized gross cost savings by year-end, far exceeding the original target of $115 million in net synergies within 24 months. On the product side, the company disclosed progress on its AI-native adaptive skills platform, which it plans to make broadly available to enterprise customers in the first half of 2027.

In addition, operating metrics disclosed in the earnings report show that the enrollment rate for generative AI-related courses rose from 25 per minute in 2025 to more than 40 per minute in 2026, and enrollments in critical thinking courses grew by more than 100% year over year.

01 Financial improvement and revenue pressure coexist

Coursera's Q3 revenue was $377 million, adjusted EBITDA was nearly $62 million, and margin was about 16%, expanding by more than 800 basis points year over year and doubling.

But standardized revenue declined by about 3% year over year, with the consumer business especially weak.

Greg Hart said at the outset: "We are strengthening our financial position as promised. Reported adjusted EBITDA margin expanded by more than 800 basis points year over year, doubling to about 16%. At the same time, our revenue trajectory, especially in the consumer business, remains below our growth targets. We continue to focus on what we can control, and that starts with integration. Since May, we have streamlined the combined organization, integrated our go-to-market teams, and aligned around a unified product vision and roadmap."

Management clearly focused on "what we can control"—integration and costs. Greg Hart said: "Integration is the foundation, not the destination. Its purpose is to create operating discipline, focus, and investment capacity to build the next step. Improving our growth trajectory requires disciplined execution and products that create significantly more value for customers."

On May 11 this year, the merger of Coursera and Udemy officially closed. The company is now advancing the post-merger ecosystem, aiming to achieve more than $150 million in annualized gross cost savings by year-end, supporting the target of $115 million in annualized net synergies, with the merger process also greatly ahead of schedule.

The integration actions include: streamlining the combined organization, eliminating duplication of overlapping functions, integrating the go-to-market teams in July, and unifying global customer coverage.

This is the quarter's strongest execution result. Greg Hart said: "We now expect to achieve more than $150 million in annualized gross cost savings by year-end. These savings support the transaction's target of $115 million in annualized net synergies, and we are on track to achieve that goal just eight months after the transaction closes, compared with the original plan of 24 months."

02 Strategic core: AI-native skills platform Project Helix

Greg Hart positioned Project Helix as the "number one priority," saying: "Absolutely, Project Helix is the future of the business and the future of both platforms."

Greg Hart introduced the product logic of Project Helix, including the following functions:

First, skills intelligence: translating business goals expressed in natural language into the required skills and development paths;

Second, personalized applied learning: continuously adapting to learners' roles, goals, progress, and organizational context, bringing learning into the workflow;

Third, Proof: scored assessments, hands-on practice, and recognized credentials, providing evidence of "what one can do."

"Project Helix aims to prove what a person can do, not just what they have completed," Greg Hart said.

In his view, connecting trusted learning, relevant context, and trusted proof at enterprise scale will be what differentiates Helix.

In September this year, after the merger of Coursera and Udemy, the company held an annual flagship customer conference, FWD Virtual, initially mainly in an online virtual format and later extended to regional events in various locations. Project Helix was previewed to customers for the first time at the conference.

Next, it will proceed in stages, co-creating with selected design partners. The goal is broad enterprise availability in the first half of 2027, with consumers to follow.

Explicitly does not assume any material Helix revenue contribution in 2027: "We will not assume Helix-driven growth contribution before the evidence supports it."

03 Demand-side logic: AI raises the bar for learning, but does not eliminate the demand for learning

In Greg Hart's view, AI raises the bar for learning rather than eliminating the demand for learning—it makes "getting answers" cheap, but makes "building verifiable capabilities" scarce.

He builds his argument with two sets of data:

(1) The gap between investment and talent readiness, "Goldman Sachs estimates that global AI-related investment will reach about $1 trillion in 2026... Deloitte found that 43% of business leaders say their organizations are well prepared in terms of AI technology infrastructure, while only 20% say the same about talent, a gap of more than two to one."

(2) Empirical evidence that chatbots harm learning, "Using a general-purpose chat-style interface improved practice exercise performance by 48%. But when tested without AI, these students performed 17% worse than the control group. The encouraging news is that tutors designed with learning safeguards largely mitigated this negative effect."

From this he draws his core judgment: "If the value of online learning were merely access to information, AI would make that value easier to replicate. The enduring value lies in trusted expertise, effective practice and feedback, and trusted proof that people can do the job."

At the same time, he also cited the 2026 Global Skills Report (eighth edition, the first based on combined Coursera + Udemy ecosystem data), pointing out that learner behavior is undergoing quantifiable structural changes—learners are not choosing between technical AI skills and human skills; they are building both at the same time.

This is reflected in:

First, demand for GenAI learning continues to accelerate. "On our platform, enrollment in GenAI courses increased from 25 per minute in 2025 to more than 40 per minute in 2026."

Second, human skills are heating up in tandem; meanwhile, enrollment in critical thinking courses grew more than 100% year over year. The growth rate is even higher than that of GenAI course enrollments, showing that learners have not abandoned "soft skills" because of the rise of AI.

Against this backdrop, Coursera's partners continue to expand, with new partners including Harvard Business School Online (leadership and business strategy courses), GitHub (courses related to agentic AI and modern software development), and others, covering both ends of "technical AI" and "human skills."

04 Ecosystem: 310 million learners, nearly 12,000 enterprise customers, dual-instructor system

The combined ecosystem now covers more than 310 million learners, serving more than 1.6 million paid subscribers and nearly 12,000 enterprise customers.

But the real moat is not these aggregate numbers, but two complementary instructor systems: Udemy brings a global community of nearly 100,000 instructors, whose value lies in "responsiveness"—when tools, roles, and skills change rapidly, practitioners can quickly turn the latest practices into courses; Coursera, meanwhile, relies on deep relationships with 425+ universities, leading companies, and professional institutions to provide trusted teaching and recognized certificates.

In Greg's words, after the merger, the two together form a "broader, more responsive foundation": Udemy solves "fast," Coursera solves "trusted," and what Helix needs to do is orchestrate the two around the same learning goal.

The ecosystem expansion in Q3 is precisely evidence that this complementary structure is beginning to produce chemistry. At the FWD conference, the company announced at once 16 new university and industry partners and 25 new professional certificates, and deliberately spanned a wide range of fields: Harvard Business School Online adds leadership and business strategy, GitHub adds agentic AI and modern software development, and the American Medical Association and Taipei Medical University reach into healthcare and life sciences...

Not betting on a single track is itself accumulating cross-industry authoritative supply for Helix's skills graph.

Anthropic nearly doubled its Coursera catalog this quarter, adding 10 courses and planning monthly updates, turning the "course catalog" into a dynamic asset that rolls out updates as AI evolves.

Two "cross-platform experiments" are even more intriguing, such as Udemy instructor Dr. Ryan Ahmed adapting a series of Claude courses into a Coursera specialization, becoming one of the most successful new launches of 2026...

This is also what Helix is doing, namely leveraging expertise across platforms to integrate content from different experts around learners' needs.

The supply side itself is also participating in the design. At the first Partner and Instructor Forum in Madrid, more than 80 leaders from universities, industry, and the instructor community worked directly with the team on the future of AI-native learning, and several have joined the design work.

The logic behind this is that if Helix is to transform expert knowledge into adaptive learning, practice, assessment, and proof, it must involve knowledge providers from the very beginning.

Taken together, what management wants to convey is a "compounding" structure: trusted institutional expertise, practitioners' responsiveness, broad distribution capabilities, plus Helix's increasingly intelligent orchestration—the combination of these four will over time make the ecosystem more valuable to every participant.

05 Financial details and outlook

Q3 key numbers