Incumbent observability tool limits visibility into Forbes’ stack
Forbes’ mission is to give people the knowledge, resources, inspiration, and connections they need to achieve success. The Forbes brand today reaches more than 140 million people worldwide through its trusted journalism, signature LIVE and Forbes Virtual events, custom marketing programs, and 43 licensed local editions in 69 countries. Forbes Media’s brand extensions include real estate, education, and financial services license agreements. Forbes.com plays a key role in delivering content and engaging with potential customers, reinforcing brand awareness, and providing information about products and services.
JD Weiner, director of DevOps, and Sameer Patwardhan, senior vice president of technology, are tasked with keeping the Forbes website available and performant. The company is a heavy user of Google Cloud and Google Kubernetes Engine (GKE) in particular. “From a developer productivity perspective, GKE is very conducive to the way we deploy our workloads,” explains Patwardhan.
To ensure its website performs as expected, the Forbes team needs consistent, cross-stack visibility into their applications and infrastructure. The company had an existing observability solution, but its per-seat pricing meant not all teams had access to observability data. Observability data was siloed and the solution was challenging to set up and navigate. As a result, troubleshooting was a lengthy, cumbersome process. “Having [the data] siloed was a huge headache,” says Weiner. “I know how a lot of our stack works, but I’m not an expert in certain aspects of it, and I wasn’t able to expose this information to the people who were."
Expanding access to observability data with Datadog
Forbes sought a new tool that would enable it to break down these silos, and Datadog was a natural fit. With Datadog, Forbes was able to expand access to observability data from a select group of operations staff to its entire technology organization. Without the constraint of per-seat pricing, the company went from 5 users with their old tool to 88 users—its entire technology organization—with Datadog.
Now, engineers can self-serve and use Datadog to see which services are taking longer to complete, perform due diligence around performance, and deliver superior software at the delivery phase. “One of our vice presidents told me he loves this tool because he’s able to search easily and diagnose problems himself,” says Patwardhan. “He no longer has to go through the operations team to get help.”
Making Google Cloud observability simple
Another key factor in Forbes’ selection of Datadog was its ease of use. Datadog comes with 800+ out-of-the-box integrations, including 35+ with Google Cloud services. “There are a ton of them, and they’re simple to set up. The documentation is very clear,” says Weiner.
With the help of these integrations, the Forbes team was able to set up Datadog in less than a day—a quarter of the time their previous solution required.
Weiner and Patwardhan have also received feedback from engineers that it’s now easier to analyze data and create the views they need, including dashboards. “Datadog’s intuitive user experience has simplified the process of optimizing applications and infrastructure,” says Weiner. “It’s a great feeling when I hear somebody else say, ‘Look at this data,’ and provide a link to the dashboard they’re looking at.”
This ease of use has also helped Forbes’ efforts to expand observability across its technology organization. Patwardhan estimates it takes only minutes to onboard a new user on the Datadog platform, due in large part to its intuitive user experience. “Datadog is easy to use because there’s no query language required,” says Weiner. “The whole platform is much more modern and flexible. It’s a lot easier to work with.”
“Datadog’s intuitive user experience has simplified the process of optimizing applications and infrastructure.”
JD Weiner
Director of DevOps, Forbes
Delivering an optimized, performant user experience
By making observability data easily accessible to its engineers, Forbes has achieved performance and optimization improvements across its Google Cloud stack, including reducing homepage load time by 37 percent.
In another notable example, Forbes’s DevOps team discovered that some of their GKE nodes were scaling to the maximum number of pods. They suspected the scaling was unnecessary, and with Continuous Profiler, they were able to identify that the cause was a specific problematic Java method. The engineering team was then able to reference these findings to identify what parts of their codebase would need refactoring. Weiner estimates that by making this change, they will be able to eliminate unnecessary scaling and bring GKE usage down by 65 percent.
This is just one piece of broader GKE optimization efforts that have reduced Forbes’ GKE costs by 33 percent to date. Weiner cites the interoperability of Datadog’s various products—driven by the auto-correlation of its telemetry data—as indispensable in enabling these efforts. “Being able to easily switch between APM and Infrastructure views of our GKE clusters has been immensely helpful,” he says.
Growing their business with confidence
Forbes is now better equipped to grow and deliver the best possible experience for millions of readers and website visitors. The company continues to innovate and introduce new capabilities, and Datadog’s growing number of integrations ensures it maintains constant visibility into its stack. “As we release new features like AI search powered by Vertex AI, we know Datadog will be there to ensure that our AI capabilities are as performant and efficient as the rest of our stack,” says Patwardhan.
No matter how Forbes’ stack evolves, Datadog ensures that observability data is democratized, making troubleshooting and optimization full-team efforts. As Weiner summarizes, “Datadog gives me—and now, our entire technology organization—invaluable insight into how our site and the systems that support it are performing to ensure they’re available, performant, and reasonably priced.”
