Docebo ARR Reacceleration Hides Cleaner Enterprise Engine

Docebo reported annual recurring revenue of $255.1 million as of June 30, 2026, up 9.5% year over year, the second straight quarter of re-acceleration after growth had slowed to 8.4% in fiscal 2025. Total revenue hit $68.7 million, up 13%, and management raised full-year revenue guidance by roughly $3.5 million while holding adjusted EBITDA targets steady to fund AI and healthcare bets.

The Toronto-based learning software company framed the quarter as proof that its shift to an AI Workforce Readiness Platform and larger enterprise accounts is taking hold. The stock’s modest premarket move suggested investors want more proof the higher-quality revenue mix can outrun lingering cash-flow noise and mid-teens headline growth.

The Quarter’s Scoreboard

Subscription revenue reached $63.8 million, up 12% including about one point of foreign-exchange help. Adjusted EBITDA climbed to $11.2 million, a 16.4% margin versus 15.2% a year earlier. Adjusted net income came in at $9.4 million, or $0.37 basic and $0.35 diluted per share. Average contract value jumped 27% to $74,800.

Docebo published its full second-quarter results and raised outlook on August 7. Full-year subscription revenue is now guided to $255.5-$257.5 million and total revenue to $274.5-$276.5 million. Q3 total revenue is pegged at $69.5-$69.7 million with adjusted EBITDA of $15.9-$16.1 million, implying a sharp sequential margin step-up.

Why the Core Growth Number Matters More

Reported ARR growth of 9.5% still carries the long wind-down of Docebo’s largest OEM customer, widely understood to be Dayforce. That relationship now represents just 2.5% of ARR, down from 8.4% a year ago. Strip out the OEM, acquired ARR, and a $0.4 million FX hit, and underlying ARR rose approximately 13.9%.

That cleaner print is the second-order story. The company spent years absorbing the OEM drag and an earlier AWS-related concentration. With those headwinds nearly lapped, the direct and partner-led enterprise engine is finally visible. Management had labeled 2026 “the year of the enterprise.” Two consecutive re-acceleration quarters give the label early data.

• OEM concentration risk falling fast, largest customer now 2.5% of ARR.
• Multi-year contracts dominate, presentation materials showed 93% of 2025 ARR additions came from multi-year deals.
• Larger customers stickier, three-year average net dollar retention ran nine points higher for accounts above $100,000 ARR than for those below $50,000.
• Customers above $100k ARR tripled from 171 in fiscal 2021 to 524 in fiscal 2025 while the smallest cohort shrank.

Subscription revenue now accounts for 93% of the quarter’s total. Gross margin slipped to 79.4% from 80.9%, partly because professional services jumped 31% to $4.8 million. That spike sits in tension with earlier comments favoring system-integrator partners for implementations, and it is one reason free cash flow compressed.

Enterprise Deals That Replaced Legacy Stacks

New logos in the quarter were heavy on complex, multi-audience deployments that bundle learning, skills intelligence, and external training.

• A global network-infrastructure and telecom leader chose Docebo plus 365Talents and a partner to unify internal training, external certifications, e-commerce, and skills-based career paths, replacing multiple legacy systems.
• The world’s largest privately owned security-services firm selected Docebo for 130,000 employees plus a multilingual customer-facing academy.
• A global automotive-safety company with more than 70,000 employees picked Docebo and 365Talents to replace siloed competency systems and Excel tracking with unified skills architecture and a talent marketplace.
• A leading global consulting firm chose the platform for worldwide internal learning, citing AI Roleplay, MCP, and Companion capabilities.
• A FedRAMP win with a private-sector energy company extended government-grade security into regulated commercial use; public-sector expansions included Kentucky, Indiana Public Retirement System, and Mississippi.

Partners sit in roughly 80% of the enterprise pipeline, whether co-sell or implementation. Deloitte, Accenture, NIIT, AWS, Carahsoft, and EY appear repeatedly in the go-to-market story. External and hybrid use cases already drive about 60% of ARR in presentation materials and carry higher win rates and ACVs.

The AI Layer Is No Longer a Slide

Docebo’s deck positions the company as the largest public pure-play in corporate learning and skills, now operating as an AI Workforce Readiness Platform across external training, internal learning, and skills intelligence. The architecture puts Docebo AI as an intelligence layer above AgentHub orchestration, with pillars for Skills Intelligence (via 365Talents), Learning Management (AI content creation, roleplay, and more than 20 capabilities), and Enterprise Knowledge.

The acquisition of 365Talents skills intelligence is six months old and already showing faster pipeline than internal targets. CEO Alessio Artuffo told analysts the pipe targets “we’ve blown that up.” AgentHub and Enterprise Knowledge are slated for general availability in early fall 2026. Custom agent workflows aimed at quick-service restaurants, healthcare, and financial services are next, supported by newly hired foundational field-deployed engineers who will first sit in R&D before monetization paths (credits or fixed price) appear.

Q2 was another milestone quarter for Docebo as disciplined execution and long-term investment continued to strengthen our position with enterprise customers around the world. As organizations transition from AI experimentation to enterprise-scale workforce transformation, they are increasingly choosing Docebo as their trusted partner.

Artuffo made that statement in the earnings release. On the call he broke the re-acceleration into three vectors: multi-year execution investments, product (organic features plus 365Talents and Zive), and the partner motion that now dominates large deals.

The AgentHub and Enterprise Knowledge platform launch turns skills detection, learning delivery, and knowledge connectors into a closed loop rather than bolted-on modules. Healthcare is the first dedicated vertical push: roughly a $3 billion slice of the corporate learning market, with Docebo already holding about $10 million in related ARR and planning content-network and partner investments over the next 12-24 months before deeper life-sciences work.

Cash, Debt, and the Buyback Signal

Cash stood at roughly $45.7 million against $88 million in borrowings at quarter-end, leaving about $42-45 million of net debt. Free cash flow fell to $3.1 million from $11.4 million a year earlier as operating cash flow turned negative $3.1 million and working capital moved to a $30.2 million deficit. Management had flagged that Q1’s elevated free-cash-flow margin included temporary working-capital benefits that would reverse.

Capital allocation is clear for now. Share repurchases sit at the top of the list. The company completed a substantial issuer bid earlier in 2026 and announced another $70 million bid at $20.40 per share in July, funded partly by the expanded credit facility. CFO Brandon Farber has described the stock as undervalued at recent prices and said buybacks remain preferred over further M&A while the shares trade at what management views as a discount. Future deals will stay selective; Docebo still calls itself “very much an organic growth shop.”

The balance-sheet shift from net cash to modest leverage funds both the return of capital and the two recent acquisitions. It also introduces interest and leverage questions that did not exist a year ago. Gross margin pressure from professional services and the cash-flow swing give skeptics concrete numbers even as adjusted EBITDA expands.

Second-Half Proof Points Already on the Calendar

Management’s confidence for the back half rests on pipeline quality, especially deals above $500,000, continued partner contribution, international strength, and the final laps of OEM drag. Q3 is seasonally important for federal and broader government work. AgentHub and Enterprise Knowledge going generally available in early fall will test whether the AI story converts into attach rates and higher ACV on new logos.

What we know

• Underlying ARR growth near 14% after adjusting for the fading OEM, acquisitions, and FX.
• ACV up 27% and multi-year contracts now the norm for new ARR.
• Full-year revenue raised; adjusted EBITDA held to fund product and vertical build-out.
• Partners involved in ~80% of enterprise pipeline; external use cases already ~60% of ARR.

What remains open

• Whether free cash flow normalizes back toward mid-teens margins after the working-capital swing.
• Attach and monetization rates for AgentHub once it reaches GA.
• Speed of healthcare vertical returns versus the multi-year life-sciences plan.
• How quickly professional-services mix reverts or gets absorbed by partners without further margin drag.

Presentation materials still target a 10-15% subscription growth band, R&D at 15-17% of revenue, sales and marketing at 26-28%, and G&A at 9-11%, with an explicit Rule of 40 ambition that puts growth first while scaling free-cash-flow contribution. Customer proof points already in the deck (Booking.com administrative hours cut, Brooks faster onboarding, Zoom training millions of customers) show the efficiency story that large buyers buy.

The second-order shift is already measurable in contract size, customer mix, and the near-disappearance of the biggest concentration overhang. Whether that quality compounds into sustained double-digit organic acceleration and cleaner cash conversion is the test the next two quarters will grade. For now the numbers show a company that is no longer the same mid-market LMS it was when the OEM drag began.