Stack traces can point out where a problem occurs, but they often fail to explain why. Logs fill that gap, providing crucial context that can save developers hours of frustration. The author shares a personal experience with a bug in a Next.js application called WebVitals, which analyzes web performance data using an AI backend. The issue arose when requests from Firefox and Safari were incorrectly identified as coming from bots, while Chrome requests went through without a hitch. To troubleshoot, the author added logging to capture the user agent string. When checking the logs, it became clear that the AI SDK was sending its own user agent, "ai-sdk," instead of the browser's. This misidentification triggered the Vercel bot protection, leading to access denial. The logs allowed for quick identification of the issue, which was tied to specific browser behaviors, ultimately leading to a simple fix: adjusting the firewall settings to bypass bot protection for requests containing "ai-sdk." The article emphasizes the importance of logs in debugging. Unlike errors, which only indicate failure points, logs provide insight into the data flowing through the system at various stages. The author highlights the benefits of high-cardinality attributes, which allow for flexible searching and filtering of log data. Connecting logs with broader trace data in Sentry offers a comprehensive view, reinforcing the idea that understanding the context of data is key to effective debugging.