A GigaOm Survey: Driving Better Outcomes with APM and Observability

Executive Summary

The world is going cloud native, both in terms of infrastructure and how applications are built. While this brings increased flexibility and potentially lower costs, organizations have found they need to address the challenges of developing and deploying cloud-based applications.

This report reviews the state of play for organizations considered actively cloud native. We examined where organizations are on their DevOps journey, and the role of Application Performance Management (APM) and/or observability tooling in this context. A particular goal of the research was to look for how deployment of such solutions had an impact on addressing the challenges and delivering on the benefits.

We surveyed 358 respondents at self-identifying cloud-native organizations with over 1,000 employees who were located in Europe and North America. Three quarters (76%) of our sample had already actively deployed some aspect of an APM and/or observability solution.

In doing so, we found that complexity is a central challenge. While this cannot be resolved entirely, tooling and practices go a long way to addressing its consequences–diagnosing problems more quickly, delivering higher quality, and becoming more productive.

Key takeaways were as follows:

• A larger proportion of the biggest companies (87%) were more likely to have actively deployed an APM and/or observability solution.
• Customer experience appears as the top justification for going cloud native (with 64% of respondents seeing it as a primary driver), followed by security (57%). The next most important criteria were scale and speed.
• Neither costs nor market trends are as important as a strategic driver (47% and 41%).
• Cloud-native applications are primarily aimed at data analytics (68%), followed by customer-facing applications (58%). Of course, the two go hand in hand–offering services to customers, then understanding how they use them.
• Application complexity is seen by respondents as by far the greatest challenge to cloud-native environments (59%). This is more of a challenge for larger than smaller companies and those less advanced on their DevOps journeys.
• 73% of respondents reported that DevOps initiatives were driven by IT, but that’s still a significant proportion (27%) driven by the business. This increases for the largest and smaller groups but not for mid-sized groups.
• Looking at which areas of the DevOps toolchain are covered, at the bottom of the list (at 51%) is operational management, reinforcing the point that DevOps is still more about Dev than Ops for many organizations.
• DevOps adoption is a clear journey, with challenges to be overcome along the way. Organizations more advanced on their DevOps journeys see the most significant benefits as software quality (70%) and developer productivity (67%), indicating the importance of achieving innovation at scale. Reduced costs is the least significant factor (46%).
• Unsurprisingly, direct improvements to application performance is the first driver to APM and observability, according to 66% of respondents. This is closely followed by the need to diagnose problems (59%) more quickly. This plays against application complexity being an issue.
• Consequently, it makes sense that real time performance measurement is seen as the most important capability of the toolset (62%).
• In terms of APM and observability features, a significant number (67%) have deployed (relatively new) database performance measures. User experience comes next, followed by lower-level measures–apart from business and retail metrics, which have been deployed by the fewest organizations (36%).
• APM and observability adopters deployed more broadly have seen significant additional benefits. For example, the sub-group of respondents that have deployed more advanced features have seen an 18-25% increased impact in all but managing costs.
• Wrapping up, the two most important criteria to further improve cloud-native application delivery are to improve performance tooling (58%) and to bring teams onto a single toolchain (57%).

Overall, we found that organizations further along the path reaped rewards. While cost controls matter at various stages of the journey, they are only one part of the whole, driving towards better business value and engineering outcomes. The most advanced, confident organizations think about what they can get out of their technology investments and not how much they can save along the way.

There are concrete cost and security advantages to implementing APM and observability, driving immediate returns. As such, businesses can start with performance tooling to gain a return on initial investment; plus, the more tools are embraced, the more significant the benefits. However, the benefits go much further. Ultimately, performance management is about doing more and better rather than doing the same for less money, as a growth play that exceeds expectations, not an efficiency play.