Koji Ikeda

Olivier argued more automation delivers more value, not less, framing the future of observability as fixing and preventing issues, and said customers who use Bits AI deploy more products and generate more usage, not a zero-sum outcome.

Pomel's view is unchanged: taking observability in-house remains a cultural rather than economic decision. Even hyperscaler teams with unlimited tooling and money choose Datadog for direct path to problem-solving.

Pomel confirmed broad-based growth largely not GPU-related. Obstler added it spans enterprise and SMB (even excluding AI companies) and across all spending ranges, a truly broad demand improvement.

Pomel confirmed expectation for more security M&A, noting both table-stakes features and AI-driven disruption create opportunities for acquisitions.

Pomel noted high retention, strong product engagement, and that churned customers often return. Short-term drops during renegotiations are expected but long-term growth follows.

Obstler confirmed FX exposure is minimal as Datadog is largely USD-based, with local billing not significant.

Pomel highlighted GPU management, IoT/on-prem fleets, and network monitoring as potential growth drivers, noting still-small market share.

Janesh confirmed multiyear-to-annual billing transition dollar impact is done but growth rates will normalize. General rule of thumb: FCF growth correlates with underlying non-GAAP net income growth, with discrete items called out separately.

Andrew said robust API usage with monetization coming for heavy machine-based usage. Most customers won't be impacted, but those running 24/7 compute on APIs will see consumption-based billing. MCPs and APIs are another monetization lever going forward.

Andrew pointed to regulatory simplification and industrialized construction (prefabrication) as future unlocks. Current backlog already exists, customers need productivity technology now.

Janesh said NRR improvement largely reflects Q2 strength and NTM mechanical effects. Underlying NRR within 100-110% range, bouncing around quarter to quarter.

Sridhar declined to share exact uplift figures but argued breadth and depth of use cases prove durability rather than pull-forward.

Sridhar cited hard-to-replicate infrastructure such as role- and row-level access control at scale, world-class replication and disaster recovery; Christian added that customers have already configured governance, security and identity in Snowflake, so AI amplifies that rather than forcing them to rebuild it elsewhere.

Brian said there were no air pockets, citing good sales execution and RPO reflecting customer business outcomes.

Khozema said both, with interactions mostly human for the next couple of years before shifting toward human-to-agent volume.

Management cited market leadership, premium pricing, multichannel reach and 4,800 interconnections across 180 plus countries with compliance complexity as its moat.

Aidan held gross profit growth similar to organic revenue, citing mix: higher-margin voice is accelerating but messaging is almost 58% of revenue.

Mike argued context is hard to build and a unique differentiator, citing quality gains in chat, semantic code indexing, and skill inference. He said most organizations will run three-to-five knowledge graphs and Atlassian opened the Teamwork Graph via MCP server and CLI to interoperate with any agentic harness.

Mike clarified 5% is differential vs non-AI cohort, not absolute rate. Cogen cohort companies are cutting-edge and growth-oriented. More code creation drives more architecture complexity, more management overhead, more security/compliance needs, all generating more Jira issues and collaboration.

Ashim pointed to net new ARR stabilizing in an 80-20/70-30 logo-vs-expansion split, driven by AI and agentic pulling through the platform.