Scaling global monitoring to match rapid market expansion

Booksy has transformed the way beauty and wellness professionals connect with their clients. Today, it operates one of the largest appointment scheduling and business management platforms in the industry. With more that 400,000 service providers and nearly 40 million consumers across 20 countries, Booksy’s platform processes millions of appointments monthly through its mobile and web applications.   As Booksy continued its rapid global expansion, it faced mounting challenges in maintaining consistent performance and reliability across its diverse tech stack. With three primary platforms—Android, iOS, and web—and mobile usage dominating customer interactions, the need for comprehensive monitoring became increasingly critical. “As we expanded into new regions and countries, we needed to ensure both rapid time to market and superior user experience,” says Adrian Jarczyński, SRE Manager at Booksy. “That meant having complete visibility into our platform’s performance and potential issues.”

With its previous observability tools, backend teams didn’t have access to frontend errors, which led to siloed communication. At the same time, tool sprawl required teams to context switch and manually correlate data across multiple tools. The SRE team needed to reduce the burden of maintaining their in-house monitoring solutions while building a more holistic view of their observability landscape. Most critically, they needed to improve application availability and performance monitoring across their expanding global footprint.

Unifying observability for enhanced user experience

The Booksy team had been using Datadog Application Performance Monitoring (APM) for backend application performance monitoring for four years, but were using another product for frontend monitoring. The SRE team recognized an opportunity to consolidate observability and decided to expand their Datadog implementation to include Real User Monitoring (RUM) and Error Tracking, creating a single point of truth to correlate all metrics.   “With Datadog APM and RUM, we now have end-to-end observability, which is crucial for our fullstack teams,” says Jarczyńnski. “The previous solution was way less flexible—with Datadog, we can correlate all metrics and have everything in one place to begin troubleshooting without needing to switch tools. That’s been a huge improvement.”

A key factor in Booksy’s decision was Datadog’s RUM without Limits model, which enables the company to capture data from 100 percent of user sessions yet retain only the most critical sessions—enabling them to reduce storage costs.

Today, Booksy uses more than 50 custom RUM-based metrics, along with dashboards and monitors, to continuously track and monitor their applications. RUM is the primary source for defining and optimizing critical user journeys from the user’s perspective. Booksy has created individual RUM applications for different platforms (iOS, Android, and web) and different user types (customers and providers). Individual teams have also created RUM custom metrics to ensure long-term access to unfiltered data and to create tailored monitors based on key business logic requirements.

The SRE team uses RUM daily to address previously undetectable challenges. For instance, before RUM, they were unaware of the prevalence and user impact of random iOS user logouts. By using RUM, they were able to decrease these occurrences from 100 cases per day to zero. “RUM allowed us to investigate and solve a baffling problem related to user logout during the startup process of some of our apps,” says Zbigniew Kamiński, head of SRE. “This would be virtually impossible without a central repository of sessions, logs, events, and metrics.”

Improving error detection, user experience, and resolution times

By centralizing observability across teams, Booksy has created a single source of truth for monitoring its complex global infrastructure, improving efficiency, minimizing downtime, and optimizing platform performance. It has also enabled Booksy to scale more effectively by using Datadog metrics to inform infrastructure decisions and optimize resource allocation. As a result, the company’s deployment frequency has increased from once weekly to 5–6 deployments per week.   In addition, RUM now provides a unified language within the organization so everyone can access and use the same relevant data. For example, C-suite executives and managers use high-level SLOs and metrics to assess business impact, while developers can dive deeper into the same data to investigate specific issues by examining errors, traces, logs, and events. “Everyone is now using the same language regardless of technical expertise,” says Jarczynski.

Perhaps most significantly, the Datadog platform has enabled Booksy to make more data-driven decisions while maintaining rapid growth across global markets and ensuring they’re bringing value to their customers. “Using SLOs and Error Tracking allowed us to start quantifying the user experience and we are already seeing improvements in these areas because teams now know which issues are really crucial,” says Kamiński. “For example, by using RUM, we have managed to isolate and eliminate one of the key reasons for erroneous logouts during app startup, removing thousands of such errors per day.”

“Organizational details and approaches change, but Datadog observability stays the same,” adds Jarczyński. “Without a comprehensive observability platform, our organization would have a very limited overview of the performance of our complex dispersed cloud infrastructure. It would be exceedingly difficult to achieve this with self-hosted or less developed solutions.”