We made a decision to utilize New Relic to the fullest extent. And we haven't had any regrets.
—Tor H. Wiig, IT Service Monitoring Manager, Elmera Group
In 2021, Elmera Group's consumer-facing apps, Fjordkraft and AllRate, were scoring around 2.0 on the App Store and Google Play. For one of Norway's leading energy companies, whose apps give customers real-time visibility into volatile electricity prices, that score wasn't just a technical problem. It was a direct reflection of customer satisfaction, retention, and trust. The engineering team had limited visibility into why.
A new approach to monitoring: Outsourced to insourced
Prior to 2021, Elmera's infrastructure monitoring was outsourced. Nothing was monitoring the application layer until 2014 and, even then, only an extremely limited portion of systems portfolio were covered. Root cause analysis was entirely reactive: when unplanned downtime happened, it was often customers who alerted the team through customer service first. "This took a long time to resolve because we didn't know where to start," says Tor H. Wiig, IT Service Monitoring Manager, Elmera Group. Major issues involved many people and considerable time before a cause was identified.
When the team first evaluated New Relic, they were skeptical. With around 10 developers at the time, things felt like they were working well enough. That changed when the team saw distributed tracing. For the first time, engineers could follow a request end to end—seeing not just that a request took three seconds, but exactly which part took one second, which took half a second, and precisely where to make a change that would actually make a difference. According to Mats Fugelsnes, Head of Engineering, that capability is what turned skepticism into conviction.
Tor's approach to rolling out New Relic was not to force the change but to show developers what was in it for them. Once they saw the value it added to how they worked, it became a no-brainer. "Everything worked like a charm," says Tor.
Today, Elmera monitors around 160 microservices and 36 different apps through New Relic. "Now, New Relic is our main and central platform for monitoring services and applications at Elmera Group," says Tor.
Fixing and maintaining the app score
Getting both apps from a score of 2.0 to 4.3 required two things working in parallel: New Relic giving the team precise visibility into what was technically wrong, and the app teams working through those issues methodically while continuing to ship new features.
"New Relic gave us great insights into the technical part of things. We located the issues and started going through them one by one, and things got so much better," says Tor.
Within a year, both apps had reached a score of 4.3 on the App Store and Google Play. Mobile application error rates were cut by more than half, dropping from between 3% to 4% to 1.5%. Today, the error rates on mobile apps are approximately 0.7%. In May 2024, the Fjordkraft app was ranked third in Europe on the LCP Delta Leaderboard for energy insights apps.
The improvement shows in Elmera's Jira data too. Since New Relic was fully deployed, the volume of app issues tracked has consistently trended downward—including issues related to the customer-facing My Page, where customers log in to manage their energy usage and subscription. "We are back to almost zero where we started," says Tor. "It's a totally different situation now."
App issues issues trend since New Relic was fully deployed in 2021.
Deployment markers have been central to maintaining that stability. When a gradual error was building in Elmera's sales system, it was difficult to pinpoint the cause. The error rate wasn't spiking dramatically. New Relic showed the team that the build began exactly when a new release was deployed, narrowing the root cause to a null reference exception that was creating a chain reaction in the system. What could have gone undetected long enough to affect customers was caught and resolved quickly.
Elmera's customer experience dashboard, monitoring My Page activity including logins, sessions, and service purchases in real time.
When a new service is created at Elmera, it immediately gets a New Relic agent. Deviations are detected efficiently, with alerts sent automatically to the teams responsible for the relevant systems. "Developers can spend more time developing rather than correcting errors," says Tor.
Extending visibility to partners
That same observability capability extends beyond Elmera's own products. Elmera provides backend services and builds white-label apps for smaller utility companies in Norway. Those partners now receive their own New Relic mobile app dashboards—built by embedding New Relic graphs as iframes inside Confluence pages, which handles security and access control cleanly—giving them direct insight into the performance of their own customer-facing products.
Elmera's team frequently detects issues with the external services their partners rely on before the providers themselves are aware. "We tell them they have an issue and they're like, 'oh, didn't know,'" says Mats. According to Mats, this happens on a weekly basis—and it applies equally to Elmera's own systems, where errors across multiple services can be quickly traced back to a single external dependency rather than an internal fault.
Observability embedded across the organization
New Relic is now deeply embedded in Elmera's engineering culture. Around 50 to 60 developers use the platform every day, and almost 100% of all core business processes that generate income are currently monitored by it.
Four or five TV screens in the Elmera office display New Relic dashboards throughout the day. One shows the main backend services running in the background—if there are deviations, services change from yellow to red, giving the business and its leaders an at-a-glance view of system health without needing to escalate to engineering.
One of the New Relic dashboards displayed on TV screens across Elmera's office, giving business leaders an at-a-glance view of backend service health, with deviations flagged instantly from yellow to red.
The impact is felt across teams, not just in engineering. When product or support teams ask why something is slow, engineers now have an answer backed by data—immediately, without having to investigate first. Performance baselines mean conversations about whether something is actually getting worse are grounded in fact rather than perception. "We use a lot less time reproducing issues locally because we can see everything in New Relic," says Mats. "When someone asks why something is slow, we have an answer with data. Instead of always saying 'let me check,' we can just show them."
For developers, the impact is visible down to the query level. Using New Relic's database queries view, engineers can see every query a service runs, identify slow ones, optimize them, and watch the results in real time—seeing a query drop from two seconds to 20 milliseconds and knowing with certainty that the change worked. "Actually being able to see the change, and not just thinking 'we made a change and it probably works'—that's really satisfying," says Mats, relaying feedback from his development team.
New Relic's security capabilities have also become part of the daily workflow. Using .NET agents, the team now receives automatic Slack alerts when new vulnerabilities or bugs are detected, allowing developers to respond and fix issues efficiently without waiting for issues to escalate.
Beyond that, Tor monitors customer experience directly through New Relic—tracking page views, unique customer logins, app versus web sessions, recurring logins, and which additional services customers are purchasing. New Relic sits alongside Google Analytics for this kind of business data, with the two systems complementing each other. Where Google Analytics handles marketing-oriented analytics, New Relic serves as a cross-check, giving the business confidence that the numbers they are seeing are accurate across both platforms.
New Relic has also changed how Tor works day to day. "I have a lot of collaboration with developers that I didn't have before because of New Relic," he says.
Faster resolution, smarter operations
Mean time to resolution (MTTR) for critical issues dropped significantly after Elmera moved to New Relic. Critical issues that used to take up to 48 hours to resolve now take between 1 to 4 hours, sometimes even faster if the issue is within Elmera’s own domain. Overall MTTR has reduced by more than 50% as a result.
Tor also independently managed Elmera's compute consumption by implementing a sliding window in their alert conditions, draining excess consumption and keeping the budget in check without requiring additional resources.
Looking ahead, Elmera is in the process of setting up formal SLAs between the Elmera Group and its subsidiaries using New Relic service levels. Mats is also exploring New Relic's MCP capability: the ability for AI to query New Relic directly when a support ticket comes in, diagnose what went wrong, and potentially resolve cases automatically. "I think that's going to be really exciting," says Mats. "Instead of having to investigate manually, the AI can actually go into New Relic and figure out why."
What began in 2014 as a tool for monitoring a handful of new systems is now the backbone of how Elmera detects issues, resolves them, and improves continuously. "I haven't really known a world where New Relic isn't part of how we work," says Mats. "What I can see is that we have big companies around us with much less observability. New Relic is part of the reason we're where we are."
