How Recurly Manages 65 Million Active Monthly Subscribers with BugSnag’s Visibility Insights
Thousands of companies today harness the power of subscriptions and manage their business on Recurly. Founded in 2009, Recurly is a subscription and billing management platform working with businesses to boost recurring revenue and create connections with their subscribers.
Recurly leverages BugSnag to ensure that their platform works for their customers. The Recurly team says that BugSnag is one of the quickest ways to identify and solve errors. As the first indication that there may be an issue, BugSnag is a big part of Recurly's weekly release cycle.
Customers of Recurly rely on the company’s ability to provide a dependable infrastructure. And Recurly leverages BugSnag.
Stacey Snyder, Senior Director of Engineering at Recurly, has used BugSnag since joining the company in 2018. Today, he oversees the product engineering team, using BugSnag to maintain the health of Recurly’s application.
Managing over 65 million active monthly subscribers for their customers, Recurly helps businesses grow smarter and stronger. Users can react in real time, iterate, and adapt to subscriber demands while having data-driven insights provided to help boost revenue. Recurly also assists in managing subscriptions at scale or tackling problems.
Like most companies, they want to provide their customers with a stable, reliable platform that leaves little room for error.
To get to that point, however, requires knowing where – and how significant – bugs are. To help ensure they are caught and corrected, Recurly uses BugSnag. BugSnag catches the errors in real-time, and most importantly, notifies the team before a user catches them.
“We use BugSnag predominately as a risk mitigation tool to make sure things are working correctly,” said Snyder. When asked about the benefits that come to mind first, he says, “intuitive and easy to use.”
BugSnag is a full-stack monitoring solution and delivers real-time data to improve app performance. In other words, the platform works to catch bugs and errors in code. BugSnag also alerts users of the severity of the bug so they can be properly fixed or gleefully ignored.
“While we have several ways to monitor errors, BugSnag is by far the quickest and easiest way for our engineers to catch them,”
Catch of the day
Recurly deploys code weekly, and uses BugSnag to confirm their platform is working as expected.
“Every week we deploy code and occasionally we do it off-cycle,” Snyder said. “We rely heavily on BugSnag to test if the features are not working correctly. BugSnag is our first indication that something is wrong, especially right after a deploy.”
“While we have several ways to monitor errors, BugSnag is by far the quickest and easiest way for our engineers to catch them”
BugSnag also enables the team to allocate their time to fixing the bugs, not just finding them.
“If we didn’t have BugSnag, it would be a lot more manual in nature,” Snyder said.
He went on to mention the benefit of automation in this case, as many of the bugs are small, essentially meaningless, and rampant. The trick is knowing which is which.
“The use of BugSnag’s UI has been helpful. If we get five thousand snags in 10 minutes, 99% of them might be the same. The tool allows us to break down the snags, see what is happening and understand it,” said Snyder.
“Our overall stability is improved. Our uptime numbers are up year over year and overall stability is very solid,”
“That is where the UI comes in handy,” Snyder said. “We can slice and dice bugs, helping us to prioritize.”
The empowerment of knowing you’re not wasting time
When something triggers an error in Recurly, BugSnag is often the first notification the team receives that something is wrong. Since it is virtually impossible to test every existing code path with a new release, BugSnag is there to catch what the QA process can’t.
When errors are found, Snyder and his team find that the timeline view within is BugSnag very helpful.
“You can see the hills and valleys of a particular [error], which aids us in the investigation process. I can narrow down to a one-week period, or a one-hour period, or drill down to the micro level. That level of information is very powerful.”
Features you can depend on
The timeline view is not the only feature Snyder finds valuable. The places he looks for this at-a- glance status update are the tool’s Stability Center and Events Usage pages.
The Stability Center provides Snyder with a comprehensive view to better understand the stability of all Recurly’s various projects, as well as their overall health.
“Our overall stability is improved. Our uptime numbers are up year over year and overall stability is very solid,” Snyder said. “More importantly, with BugSnag we find critical issues substantially faster allowing us to either perform a rollback or push out a fix before our end users experience any issues.”
The Events Usage page showcases a graphical representation of all received errors over the last 30 days, so the team can understand the volume of errors relative to previous points in time.
By utilizing BugSnag’s robust notification system and features, Recurly’s team monitors Slack, Jira, and email to ensure they are aware of bugs with their new code, allowing the team to fix them quickly.
