Aaron Cass on ClickHouse AI revenue growth

TL;DR

Revenue growth is accelerating, with ClickHouse already reaching over 350 million ARR and aiming for a billion ARR with strong investor support. The company balances efficiency, open source adoption, and a broad product roadmap while cautioning about revenue durability due to high switching costs in AI applications. The path to scale involves expanding sales capacity and enterprise motion without sacrificing product-led growth.

Transcript

We're just getting started. This seems to be accelerating at an unprecedented pace. We haven't seen revenue growth like this in our lifetime. We went 0 1250 200 and we'll finish this year north of 500. Our guest today just got front of shirt sponsorship for Craven Cottage and for Fulham. Most importantly, welcome Aaron Cass, founder and CEO of Clic... Read More

Key Insights

• ClickHouse is described as the world’s most popular open source database known for fast queries and efficient storage.
• The biggest investor concern is the durability of revenue due to high switching costs for infrastructure software and rapid model-provider changes.
• The company prioritizes revenue growth and is willing to rein in token costs if growth remains strong.
• The founder emphasizes the need to scale sales capacity beyond a small 100-person team to compete with larger incumbents.
• Lessons from Salesforce include overestimating yearly progress but underestimating five-year outcomes, guiding the growth strategy.
• Open source adoption and AI agent workloads drive demand for the ClickHouse platform and its evolving ecosystem.
• The interview discusses balancing multiple model providers and proper model routing to optimize production code quality.
• There is a focus on a long-term roadmap for 2027 and beyond, with the belief that enterprise sales will be crucial to reach higher ARR.

Questions & Answers

Q: How does ClickHouse view its revenue durability in the AI era?

Revenue durability is seen as the single biggest risk because switching costs for infrastructure software are high, while switching costs for agentic AI applications can be low. The company monitors the durability of ARR as AI models and harness providers evolve, and it emphasizes maintaining a broad surface area with strong growth to offset potential churn. The approach is to demonstrate sustained revenue growth and manage costs as needed, ensuring a path to margin expansion while growing usage across multiple use cases.

Q: What was the founder's takeaway from Salesforce that influences ClickHouse strategy?

A key takeaway is that you can overestimate what you can achieve in one year and underestimate what you can achieve in five. This informs a long horizon for growth, where initial focus on product and engineering gives way to a layered enterprise sales motion. The lesson guides the company to balance product-led growth with strategic enterprise engagement to reach larger customers and higher ARR over time.

Q: Why is expanding the sales team important for ClickHouse?

Expanding the sales team is important because, although the company began with strong product-led growth, large data warehousing competitors rely on bigger sales and marketing efforts. A larger dedicated sales capability would enable ClickHouse to pursue enterprise-level deals and address a broader set of use cases, aligning sales motion with a mature roadmap and accelerating revenue growth toward the desired ARR target.

Q: How does ClickHouse plan to handle token spend in AI applications?

Token spend is managed by focusing on revenue growth, while token costs are expected to decline over time as efficiencies improve. The company believes it can rein in token consumption without sacrificing growth if the broad surface area of use cases continues to expand. The strategy centers on leveraging efficient AI infrastructure to drive demand and monetization through higher ARR.

Q: What is the role of model providers and harnesses in ClickHouse’s strategy?

The strategy involves using a mix of open weights models and frontier models, with consideration for optimal model routing for different tasks. While some open weights models are used for code review, ClickHouse remains cautious about production outputs. The idea is to balance independence with performance by selecting models that meet reliability and speed needs for production workloads.

Q: What lessons guide ClickHouse’s approach to enterprise sales versus PLG?

The company learns from data points like Snowflake vs Data Dog, aiming to combine PLG with enterprise sales. Initially prioritizing product and engineering, the team now sees value in layering an enterprise motion to reach large customers. The approach seeks to maintain product-led benefits while enabling large-scale enterprise adoption and higher ARR growth.

Q: What is the significance of a long-term roadmap, according to the interview?

A long-term roadmap is central to guiding investments, product development, and go-to-market strategy. The conversation notes that being behind on a roadmap would imply missed opportunities for 2027 and beyond. It emphasizes continuous feature delivery across a broad use-case surface to sustain growth and attract enterprise customers over time.

Q: How does ClickHouse view the potential for AI agents as customers?

The interview suggests that AI agents could become customers as they evolve, shifting the dynamic of how software is consumed. This implies a future where usage and value are driven by agent-based workloads, requiring a scalable and adaptable platform architecture, clear pricing, and robust integration capabilities to capture ongoing demand from AI-driven workflows.

Summary & Key Takeaways

• ClickHouse is positioned as a fast, resource efficient open source database used by major AI native companies, with ARR surpassing 350 million and intentions to grow toward 1 billion. The interview emphasizes revenue durability as a key risk due to high switching costs in infrastructure software and evolving model costs.

• The guest highlights the need to scale sales capacity and layer in enterprise sales alongside product-led growth, drawing lessons from Salesforce and Snowflake on go-to-market tension and the importance of a long-term roadmap for enterprise adoption.

• The conversation covers token spend, model routing, and the balance between open models and frontier models, stressing that sustained revenue growth and utility across a broad use case surface drive long-term value for ClickHouse.