Box, Inc. (BOX) Earnings

Management highlights

- **Financial Performance**: Exceeded guidance across all top and bottom-line metrics. Delivered five consecutive quarters of accelerating revenue growth in constant currency. Achieved a net retention rate of 106%, driven by seat expansion and price increases within the Enterprise Advanced base. - **Product Innovation & AI Strategy**: Launched new security capabilities including agent guardrails, prompt injection detection, and third-party agent oversight to secure AI workflows. Announced Model Context Protocol (MCP) integrations with Anthropic Claude, Databricks, IBM Watson X, and others, enabling agentic access to enterprise content without data movement. - **Platform Evolution**: Transitioning from static content management to an agentic workflow automation system. Enhancements include Box Automate, Box Extract for complex document processing, and improved metadata/large file support to facilitate migration from legacy on-premises ECM systems. - **Go-to-Market Execution**: Accelerating adoption of Enterprise Advanced through vertical-specific motions in financial services, life sciences, government, and legal. Expanding Forward Deployed Engineering (FDE) teams to assist customers in tuning AI agents and optimizing model selection for specific use cases. - **Strategic Partnerships**: Deepening collaborations with hyperscalers (Amazon, Google), system integrators (Slalom, DataBank), and AI labs. Key wins include a multinational investment bank upgrading to Enterprise Advanced and a major federal agency replacing legacy contract lifecycle management platforms. - **Security & Governance**: Emphasizing the need for a model-neutral platform to manage rising AI costs and ensure compliance. Highlighting robust audit logs, real-time security alerts, and identity verification to protect unstructured data against both human and agentic risks.

Guidance

- **Q3 FY2027 Revenue**: Expected to be approximately $329 million, representing ~9% YoY growth (11% constant currency). - **Full Year FY2027 Revenue**: Raised by $10 million to approximately $1.29 billion, representing 10% YoY growth (11% constant currency). - **Gross Margin**: Q3 expected at ~80.5%; Full Year expected at ~80.5% (with Q4 projected at ~80%). - **Operating Margin**: Q3 expected at ~28%; Full Year expected at ~28%. Includes an estimated 80 basis point headwind from foreign exchange. - **EPS**: Q3 expected at ~$0.39; Full Year raised to ~$1.54 (includes ~9 cent FX headwind, representing a 1-cent increase vs. prior guidance). - **Billings**: Q3 growth expected to track roughly with revenue growth (~9%). Full Year billings growth expected to align with revenue growth, including an estimated 150 basis point FX headwind.

Segment performance

Box operates as a single reportable segment focused on its Intelligent Content Management Platform. Consequently, there are no distinct product segments with separate financial reporting. The company reported total Q2 revenue of $321 million, representing a 9% year-over-year increase (11% in constant currency). Revenue contribution is categorized by customer tier rather than product line: Suite customers accounted for 69% of total revenue, an increase from 63% in the prior year period.

Risks & headwinds

- **Foreign Exchange Headwinds**: Significant FX impacts noted, including an 80 basis point headwind on operating margin and a 9-cent headwind on full-year EPS. - **Cloud Infrastructure Constraints**: Limited access to certain public cloud infrastructure components is impacting efficiency gains and gross margin performance, though constraints are expected to ease over time. - **AI Security & Governance Risks**: Enterprises face heightened risks regarding unauthorized data access by AI agents. Incidents like the OpenAI/Hugging Face breach underscore the necessity for robust guardrails, anomaly detection, and strict access controls to prevent 'rogue' agent behavior or data leakage. - **Legacy System Migration Complexity**: While demand for migrating off legacy ECM systems is growing, these processes can be long and painful, requiring significant organizational change and technical expertise to integrate AI agents with previously siloed unstructured data.

Analyst Q&A

• Q: How do token costs and open-source models impact unit economics and agentic workflows?

  A: CEO Aaron Levie explained that longer-running agents and large-scale data processing are highly token-intensive, making cost optimization critical. Box’s model-neutral agentic harness allows customers to direct workloads to the most cost-effective model—whether frontier or open-weight—based on accuracy requirements. This dynamic benefits Box by driving higher consumption volumes as customers seek to balance cost and performance, while also allowing them to leverage price reductions from AI labs like OpenAI or Gemini.

• Q: What drives the shift toward "wall-to-wall" enterprise deployments versus departmental pilots?

  A: Levie noted that as enterprises deploy diverse AI agents (e.g., ChatGPT, Cursor, Salesforce AgentForce), they require a unified source of truth for unstructured data to ensure agents operate on authoritative corporate knowledge. Fragmented data landscapes create security and governance risks. Dylan Smith added that Enterprise Advanced capabilities enable cross-departmental workflows (like contract lifecycle management), driving seat expansion as multiple departments must access the platform to collaborate securely, thereby boosting net retention.

• Q: How is Box addressing the "zero data retention" requirement amid evolving AI lab policies?

  A: Levie stated that zero data retention (ZDR) has become an industry standard, allowing enterprises to trust AI models with sensitive context. He referenced the Fable incident as a cautionary tale where lack of ZDR hindered adoption. Box ensures its GA models meet ZDR and compliance criteria, acting as a trusted layer that bridges AI breakthroughs with enterprise governance needs. He anticipates competitors like Anthropic will eventually align with ZDR standards to maintain revenue relevance.

• Q: What are the key learnings from the Forward Deployed Engineering (FDE) motion?

  A: Levie described FDEs as crucial for helping customers tune AI agents, select appropriate models, and run evaluations on their specific document types. This high-touch approach addresses the complexity of back-office and customer-facing workflows that differ from internal coding tasks. By programmatizing this expertise, Box helps customers realize productivity gains and cost savings, often leading to deeper integration and expanded usage after initial engagements demonstrate value.

• Q: Why is Q3 billings growth guidance lower than Q2 despite strong momentum?

  A: CFO Dylan Smith attributed the step-down in billings growth to inherent lumpiness and varying foreign exchange impacts compared to prior periods. He emphasized that full-year billings expectations have been raised significantly, and underlying indicators like RPO, net retention, and pipeline remain healthy. Billings growth is still expected to lead revenue growth on a constant currency basis, with the Q3 dip reflecting prudent guidance practices rather than a slowdown in demand.