Asana's CEO on turning an $800M SaaS giant into an operating system for human-AI teams
Key points
Key takeaways from a BILLIONS interview with Asana CEO Dan Rogers (September 2026):
Taking over a scaled business. Rogers joined Asana as CEO in July 2025, taking command of a work-management platformwith roughly $800M annual revenue and 18 years of history, after leading feature-management firm LaunchDarkly. Thirteen months in, he says operating margin reached a company record, the leadership team was rebuilt, $75M was spent acquiring Stack AI,andthe US government market was opened.
Listening before deciding. His playbook fora new CEO is hearing 100 customers in 100 days, triaging the company's inherited norms and decisions into keep, investigate and discard buckets, and testing ideas through what he calls iterative porpoising, moving back and forth between big-picture hypothesesand detail before committing, so he comes in with questions rather than answers.
From work management to human-agent teams. Asana now frames itself as the operating system for teams made of both humansand AI agents, serving more than 300,000 customersand 85% of the Fortune 500; Rogers argues that answers from models are not the same as coordinating action, which is where contextand a governance harness on top of the models matter.
Agentic work management. The platform ships 30 pre-built teammate agents that surface contextually inside projects, take and complete tasks, write results back into the living plan,andcan be corrected by team members, with humans approving the workand signing off at the end of each cycle.
AI is driving new revenue. AI products — the AI teammatesand AI Studio lines — now contribute 25% of Asana's net new ARR (up from 17%),and most new deals with its largest customers are AI-attached.
Request-based pricing. Instead of metering tokens or credits, Asana charges for fulfilled requests —a unit of completed agent work —and obfuscates the token complexity behind it. It also gives all users an initial allowance of five free requests to discover the value before they pay.
Models as an interchangeable layer. Rogers treats models as largely fungible and, he suggests, commoditising; Asana routes each task to the model best suited to it,and integrates with foundational assistants such as Codaand Claude, so customers can work with Asana data from inside those chat surfaces.
Buying time with Stack AI. The acquisition of no-code AI-agent builder Stack AI accelerates Asana's product roadmap; Rogers stresses that the small details of integrationand keeping the acquired team matter as much as the strategy.
Read more: BILLIONS Podcast · Asana
