Lyft

About Lyft

Lyft is a leading ride hailing marketplace offering ridesharing, bikes, scooters, and transportation services across North America. With a purpose to serve and connect, Lyft operates at an immense scale—supporting hundreds of engineers across dozens of teams and managing a highly complex infrastructure across Kubernetes, cloud providers, and additional services.

As Lyft’s cloud environment grew, so did the need for a centralized, flexible FinOps solution that could bring visibility, accuracy, and accountability to every part of the organization.

The Challenge: Cloud Cost Clarity at Scale

Managing cloud costs at Lyft meant tackling major challenges:

• Broken Attribution: Existing internal tools couldn’t align cloud resource usage with the evolving organizational structure, leading to frequent inaccuracies.
• Shared Cost Complexities: Critical platforms like Spark, Trino, and analytics events had no native way to allocate team-level or workflow costs using cloud provider tools alone. To bridge the gap, Lyft maintained its own pipeline of tools, queries, and visualizations—introducing significant maintenance overhead, technical debt, and reliance on custom logic to account for discounts.
• Visibility Gaps: Cloud billing lacked visibility, especially with delayed application of enterprise discounts, leaving teams in the dark.
• Accountability Obstacles: Without clear, team-based cost data, driving financial accountability across engineering teams was impossible.
  
  

The Solution: Centralized, Granular, and Actionable FinOps

With Finout, Lyft gained powerful new capabilities:

• Fixing Ownership with Virtual Tagging. Finout s Virtual Tags gave Lyft a powerful way to unify cost ownership across multiple platforms abstracting away technical complexity for end users and reducing cognitive overhead. While Virtual Tags didn t solve attribution alone, they enabled Lyft to present consistent tagging across environments without requiring hardcoded structures or manual updates.

  “Finout’s most powerful feature? Virtual Tags. They simplify complex cost data into a single, clear view across Kubernetes, cloud provider, and vendors.”

  Nic Baumann | Manager, TPM - Infrastructure

• Accelerating Cost Anomaly Detection. Finout’s anomaly detection engine slashed time-to-detection from weeks to just days, helping Lyft stay ahead of unexpected cost spikes and drive better operational agility. Paired with shared cost visibility, the team can now detect anomalies that previously went unnoticed.

  “The anomaly detection has been incredibly powerful. It used to take us weeks—or we wouldn’t even catch issues at all. Now, we catch them in just days.”

  Nic Baumann | Manager, TPM - Infrastructure

• Breaking Down Shared Platform Spend. Using Finout’s shared cost capabilities, Lyft could finally distribute the costs of high-usage platforms like Spark and Trino across the teams consuming them—accurately, fairly, and transparently.

  “Finout’s shared cost feature lets us break down huge platform costs across teams based on usage — something cloud providers alone simply couldn't do.”

  Nic Baumann | Manager, TPM - Infrastructure

• Streamlining Team Adoption with Tailored Dashboards. Instead of overwhelming engineers with complex billing data, Lyft created pre-built dashboards using Finout. Product teams received simple, actionable views, while power users had full access to MegaBill for advanced troubleshooting.

The Results: Visibility, Flexibility, and Control

With Finout, Lyft achieved over 96% cost attribution coverage across its infrastructure, significantly accelerated anomaly detection cycles from weeks to just days, and empowered dozens of engineering teams with self-serve cost dashboards. Finout’s shared cost capabilities enabled Lyft to gain visibility into previously untraceable spend across platforms, while custom cost pipelines allowed them to integrate non-cloud provider expenses seamlessly—laying a strong foundation for future unit economics initiatives as they advance in their FinOps maturity.

By leveraging Finout’s cost optimization and cloud spend observability capabilities, Lyft achieved a substantial reduction in its cloud provider infrastructure costs. 

And they achieved it all with cloud-agnostic flexibility, no-code onboarding, and enterprise-grade security, including SOC 2 compliance.

What’s Next: FinOps at the Unit Level

Looking ahead, Lyft is preparing to expand its use of Finout even further:

• Integrating Usage Metrics: Bringing ride growth data and other KPIs into Finout to align costs with business outcomes.
• Driving Unit Economics: Moving from overall infrastructure costs to actionable insights like cost per ride or cost per user session.
• Enhancing Anomaly Detection: Layering business metrics onto cost alerts for smarter, context-aware FinOps governance.