%201.webp){kind=small}
“All the usual APM platforms were simply too expensive for us. We were literally getting ready to spend months building and piecing together open-source tools, when one of the engineers said ‘Hey, you guys should check out groundcover’, and the rest is history.”
About Slickdeals
Slickdeals is the leading and most trusted online community for finding and sharing the best deals. The site serves millions of users and consistently ranks among the top internet forums, which means reliability and performance are mission-critical. On the technology side, Slickdeals operates with a relatively small engineering team (~60 engineers), including a platform/devops group of just 3-4 people. Recently, the company undertook a major cloud migration, moving its entire platform from legacy on-premises VMs to Amazon EKS (Kubernetes) across development, staging, and production environments. This modernization promised scalability and efficiency – but also introduced new complexity in how Slickdeals monitored and observed its countless services.
The Challenge: Outgrowing a legacy monitoring stack
As part of the migration, Slickdeals containerized its core applications and broke monolithic components into microservices. The lean platform team not only had to provision and orchestrate Kubernetes clusters, but also ensure proper monitoring for this new cloud-native stack. Historically, the company’s monitoring and logging relied on legacy tools: Nagios for basic uptime alerts, Cacti/Graphite + StatsD for custom metrics, and an ELK stack (Elasticsearch/Logstash/Kibana) for logs. These open-source solutions had been sufficient in the past, but lacked the deep visibility (like distributed tracing or container-level insights) that a microservices, Kubernetes-based architecture would require. Maintaining this patchwork of tools also consumed valuable time from the small platform team.
Facing a rapidly evolving infrastructure, Slickdeals needed to upgrade its observability approach, without overburdening its resources. The cloud migration was on a tight schedule (initially aimed at roughly a year), so the directive was to “get everything running in the cloud, and we’ll figure out how to monitor it later.” This meant that as services were moved to Kubernetes, the team temporarily leaned on minimal monitoring (some CloudWatch and ad-hoc scripts) to get by. The plan was to build a proper observability stack in-house once the migration was done: likely using Prometheus for metrics and Grafana for dashboards, tied into their existing StatsD/Graphite pipeline for continuity.
However, rolling out a full Prometheus/Grafana stack, with all the integrations and long-term storage it would entail, is a big undertaking for any team, and even more so for a lean one. Slickdeals’ engineers estimated it would take significant engineering cycles to implement and maintain. Alternatively, they considered commercial Application Performance Monitoring (APM) suites, but quickly realized those were not financially viable. “Every single major APM vendor was cost-prohibitive, there was no way we’d get that budget approved,” JT noted, referring to well-known solutions like Datadog or New Relic.
“groundcover offered so much of the things that we wanted to build, essentially everything on our wishlist, and it did so out-of-the-box”
- JT, Director of Platform Engineering, Slickdeals
Why groundcover?
During the migration, one of Slickdeals’ SREs discovered groundcover. The team decided to give it a try and deployed groundcover on one of their Kubernetes clusters for a proof-of-concept. They were immediately impressed by how quickly groundcover started delivering insights that the team actually ended the POC early and moved forward with a full adoption. groundcover had proven it could fulfill the company’s needs in a fraction of the time and effort of the DIY route.
groundcover addressed Slickdeals’ requirements on multiple fronts:
- Full-stack visibility from day one: By installing groundcover’s lightweight eBPF-based sensor as a DaemonSet in their clusters, Slickdeals instantly gained visibility into container-level metrics, application traces, and even network flow data - all without writing a single line of instrumentation code. This unified platform replaced the need for various tools (CloudWatch, custom scripts, separate tracing systems) by capturing everything in one place with minimal overhead. In minutes, the platform provided them visibility into what the team estimated would have taken months to implement on their own. This not only accelerated their observability roadmap, it also eliminated a huge maintenance burden from the small platform team.
- Unmatched cost efficiency: Unlike all other APM vendors, groundcover’s BYOC deployment model means the data never leaves Slickdeals’ own environment. This enables groundcover’s pricing structure to be completely decoupled from data volumes. Slickdeals could afford to monitor everything - every service in every environment, including dev and staging clusters - without worrying about a hefty bill. As JT noted, groundcover was a solution whose value far outweighed its price, making it an easy sell internally.
- Developer-friendly experience: groundcover’s intuitive UI and rich visualizations made it easy for engineers to explore and use. Features like the real-time network map gave instant clarity into how microservices communicate, and auto-detected issue alerts provided context (K8s pod info, relevant logs and metrics) in a single view. This meant that not only the DevOps team, but also application developers could quickly get on board. The rapid pace of feature releases from groundcover’s team, as well as highly responsive support, was an added bonus. The Slickdeals team saw new capabilities and improvements rolling out frequently, further building their confidence in the platform.
With these advantages, groundcover checked all the boxes for Slickdeals. The team realized they didn’t need to piece together and constantly maintain an open-source stack or compromise on coverage due to cost. groundcover was delivering exactly what they needed (and more), right when they needed it. It accelerated their migration project by offloading observability concerns to a ready-made solution.
“groundcover wiped out two quarters’ worth of work we had planned for two engineers - a huge win since we didn’t have to build those capabilities ourselves.”
- JT, Director of Platform Engineering, Slickdeals
The Impact
After the speedy evaluation, Slickdeals standardized on groundcover as its observability solution across the organization. They deployed groundcover on all their Kubernetes clusters - in development, staging, and production - making it the default monitoring and troubleshooting tool for every environment. This rollout even extended to individual development teams. Engineers who manage their own dev/test clusters now had groundcover at their fingertips to monitor services, without having to rely on a central ops team or multiple disparate tools. groundcover’s Kubernetes-native design and ease-of-use led to rapid internal adoption, and quickly became the go-to platform when anyone at Slickdeals needed to investigate an issue.
By implementing groundcover, Slickdeals transformed its observability stack and achieved several key results:
- Faster troubleshooting and root-cause analysis: With all telemetry (metrics, logs, traces) in one place and correlated through a single interface, teams can identify issues in a fraction of the time it used to take. No more jumping between Nagios alerts, Kibana logs, and CloudWatch metrics - a process that often slowed down investigations. Now, when an incident occurs, engineers’ first step is to open groundcover, where they can see everything from pod health to application traces instantly. For example, Slickdeals’ team leveraged groundcover’s interactive network map to debug a complex microservices issue during a loyalty service migration. They could visually spot which service connections were failing and dive into the relevant logs. The net effect is a significant reduction in Mean Time to Resolution (MTTR) for incidents, boosting Slickdeals’ uptime and user experience.
- Months of engineering effort reclaimed: groundcover didn’t just accelerate troubleshooting - it also gave back valuable developer time that would have otherwise been spent building and maintaining an observability stack. By adopting groundcover instead of pursuing a DIY Prometheus/Grafana solution, Slickdeals eliminated an estimated six months of work for its platform engineers. Those engineers were able to focus on delivering new features and improvements to Slickdeals’ product (and on the cloud migration itself) rather than reinventing monitoring tools. This productivity gain is directly tied to business value: the small team can do more with the same resources. And because groundcover continues to evolve with new capabilities, Slickdeals benefits from ongoing improvements without extra effort on their part.
- Unified platform and full environment coverage: The adoption of groundcover allowed Slickdeals to consolidate multiple monitoring systems into one. This simplification reduced operational overhead by having fewer systems to manage and integrate, and training effort as engineers only need to learn one interface. groundcover effectively replaced legacy systems like Nagios and portions of their ELK stack, as well as eliminated the need to use CloudWatch’s cumbersome interface for container metrics. All teams now speak the same observability “language” using groundcover. Moreover, Slickdeals achieved 100% observability coverage across all its services and environments. In the past, high costs or limited tooling meant certain non-production clusters or less critical services might go unmonitored or under-monitored. Now, with groundcover’s data volume-agnostic model, every service, big or small, is monitored. Developers can enable granular logging or tracing in lower environments without fear of incurring huge costs, which leads to better testing and more confidence before deploying to production. The result is a more reliable, resilient platform for Slickdeals’ end-users and a happier, empowered engineering team internally - all achieved without breaking the bank.
“Using groundcover’s network map made it so much simpler to debug. What would have taken me a week or two to hunt down, I found in a couple of hours. I was basically done in a day,”
- JT, Director of Platform Engineering, Slickdeals
Slickdeals’ journey with groundcover showcases how even a lean team can adopt and leverage state-of-the-art observability. In a short span, the company went from a fragile mix of old monitoring scripts to a modern, eBPF-powered observability platform that covers their entire Kubernetes stack. Not only did groundcover meet Slickdeals’ technical needs, but it also proved its value in business terms: reducing costs, saving time, and enabling growth. JT and his team now rely on groundcover as an everyday part of their workflow. It’s not just for firefighting incidents. They use it proactively to understand system performance, verify deployments, and gain insights that inform optimization efforts.
As JT sums up their experience, “for the value it provides, its cost is unmatched”, making it an indispensable tool as Slickdeals continues to scale its cloud infrastructure and deliver great deals to millions of users.
Sign up for Updates
Keep up with all things cloud-native observability.
