Lower costs, more visibility: Inside AuditBoard’s migration to Grafana Cloud

When Neil Laughlin joined AuditBoard in 2023, its culture impressed him right away.

“This is the most engineer-oriented infrastructure organization I’ve led,” recalled Laughlin, the vice president of site reliability engineering. “And the difference that makes in the platform capabilities is very clear.”

With that engineering-first mindset, teams needed greater visibility across environments — and the ability to experiment freely with new configurations. The company previously relied on proprietary vendor for metrics and tracing, and while the product delivered valuable insights, its costs were spiraling out of control, especially when monitoring basic but business-critical metrics. 

“I’m just trying to figure out if we need a PVC [persistent volume claim in Kubernetes] or not,’” said Puneet Kandhari, engineering manager for platforms and infrastructure at AuditBoard. “And it’s like, ‘Well, that cost me three grand.’”

 “And a conversation with your vice president afterwards,” Laughlin added, with a laugh.

Those moments encapsulated a growing tension: Engineers wanted more visibility and flexibility, but the business needed predictability and sustainability. That tension led AuditBoard on a path to rethink its approach and evaluate other observability vendors, eventually migrating to Grafana Cloud. In doing so, the company not only cut costs but also expanded their observability footprint, helping them gain more control over their data.

In addition to cutting AuditBoard’s metrics ingest from about 20 million to 9 million active series within one day of onboarding, “moving to Grafana taught me so much about what a reliable monitoring infrastructure looks like," said a senior software engineer on AuditBoard’s platform team.

‘I was apologizing to finance every month’

AuditBoard is a fast-growing SaaS company in the governance, risk, and compliance industry. With a centralized team building and running its platform, observability is vital for maintaining reliability and performance.

AuditBoard’s previous observability vendor had its strengths. In particular, its tracing capabilities had helped AuditBoard’s engineers dig into code behavior and uncover performance optimizations at a critical time for the business. But because of the cost, observability at AuditBoard was largely limited to production and debug environments. 

“Application-level metrics are much more valuable to us,” said Laughlin. The trouble was that every time the team attempted to extend their visibility into staging and QA, the previous solution’s node-based billing made it difficult to predict monthly costs. So each month brought new, unexpected spikes. 

“There was so little ability to control anything with our prior platform. We were just sucking up bills, and I was apologizing to finance every month,” Laughlin said. “It was just insurmountable in terms of cost.”

The search for a new platform

AuditBoard’s platform team briefly considered running their own observability stack, before looking at other options. Then Grafana Cloud entered the picture and immediately stood out—and not just at a technical level. It also aligned from a cultural and financial perspective.

Grafana Labs’ “big tent” philosophy — the belief that organizations should be able to choose their own observability tools and strategies and centralize their data in Grafana — aligned with the team’s preference for open source software such as Prometheus and Grafana Mimir for metrics, OpenTelemetry, and Grafana Loki for logs.

Juan Beltran, an Infrastructure and FinOps engineer at AuditBoard, noted that Grafana’s usage model also made the team’s expenses more transparent.

“Since Grafana Cloud’s usage is pretty parallel with our cloud spend and our cloud usage, it’s something that we monitor very thoroughly and try to stay aware of as much as possible,” he said.

Grafana Cloud also checked key technical boxes:

• Its native support for OpenTelemetry would allow the team to future-proof their stack.
• Open standards offered greater flexibility around data collection and integration.
• Built-in compatibility with Node Exporter would make it easy to capture essential infrastructure metrics without incurring runaway costs.

With that, the decision to migrate to Grafana Cloud was easy.

AuditBoard’s migration to Grafana Cloud

Migrations, however, don’t come easy. AuditBoard’s team faced its share of hurdles, such as cardinality explosions that come with standing up a new stack. Plus, they were learning a new product with its own set of best practices and specific configurations. 

The experience highlighted the need for significant upfront engineering investment and cultural alignment when migrating observability platforms. A high-touch partnership with the Grafana Labs Professional Services team helped AuditBoard navigate this process more confidently, accelerate troubleshooting, and build internal expertise. 

In weekly working sessions, Alex Martin, principal solutions architect at Grafana Labs, quickly sized up what the AuditBoard team already knew how to do and focused on closing the remaining gaps in migrating to Grafana Cloud. “We’ve solved way more issues with them than we would have otherwise,” Kandhari said.

It was a stark difference from working with AuditBoard’s previous observability vendor where most questions were funneled through a customer success manager and support tickets. “We didn’t get that type of engagement with our former solution,” Beltran said, citing a lack of hands-on, technical support.

“The partnership with Professional Services has been great,” echoed Geoff Schultz, Manager, Infrastructure Engineering at AuditBoard.

Once workloads began flowing into Grafana Cloud, the impact was immediate. Adaptive Metrics, a popular feature in Grafana Cloud that aggregates unused and partially used metrics into lower cardinality versions, proved to be a game changer: Within one day of the completed migration, AuditBoard cut its metrics ingestion by more than half.

Unlocking value with Grafana Cloud 

The cost benefits of Grafana Cloud were realized beyond slashing active series. Grafana Cloud also gave engineers a faster feedback loop. With Adaptive Metrics, the team could turn a metric off, watch to see if a spike disappeared and then adjust it — work that previously would have required weeks of debugging and optimization.

“We would have just had to manually configure all those drops so it’s like a time saving for engineers, which I think is a better measurement of how Grafana Cloud supports us,” Schultz said.

The teams also gained greater visibility into their production environments using Grafana Cloud’s out-of-the-box solutions, such as the prebuilt Kubernetes dashboards, billing dashboard, and cardinality management dashboard. Now, instead of having an engineer manage their clusters, they have a Grafana dashboard that helps alert on the health and efficiency of their infrastructure. 

“Once we pivoted to the Grafana-native approach, we unlocked a lot more value out of the visualizations,” Laughlin said.

By the end of 2024, the team achieved company-wide recognition for their migration to Grafana Cloud. The company’s CEO even called out the project as an example of forward-looking innovation for other teams to follow.

“We have earned trust with the leadership team in part by proactively looking for this type of opportunity and then delivering it successfully,” Laughlin said.

Fast forward to today, and AuditBoard continues to refine its observability practices at the same time it expands its adoption of Grafana Cloud across all their telemetry singals—metrics, logs, traces, and profiles. 

Grafana Cloud gives the team clearer insights into who is using the platform and why, helping them classify different usage groups to better inform future decisions about new dashboards and features. It has also allowed AuditBoard to future-proof its stack.

“One of the things that I was excited about in this migration was getting into the OTel world,” said Laughlin. “The metrics lock-in we had with our previous platform, while convenient for getting off the ground, was just the wrong collection method to be using long-term. It has been interesting learning the degree to which observability providers are now consuming OpenTelemetry as a de facto standard.”

Laughlin is also giving his engineers the skill set needed to stay flexible as innovations around AI and open standards continue to push the boundaries of observability. 

“The holy grail is something so much better in terms of efficiency and operations,” said Laughlin. With Grafana Cloud, “we’re getting maximum value per dollar and we’re spending on a platform where the skills the team is developing are useful to engineers working in this space.”

In fact, moving to Grafana Cloud has been so successful that it has allowed AuditBoard’s engineers to  to focus on engineering, which has enabled them to grow in new and unexpected ways.

As a result, said Beltran, “The sky’s the limit with how far we can go.”