At Embrace, we live and breathe mobile, and we care deeply about the end-user experience. And one thing’s always clear: mobile engineering is hard. Really hard. Between the ever-growing number of devices, operating systems, app versions, and unpredictable network conditions, developers are working in an environment full of variables they can’t always control. Throw in a complex backend system and a smattering of limited tools and…you get the picture.
So when our team in the field saw the shock on SRE, DevOps, and Platform Engineers’ faces when we explained the crazy complexity of dealing with mobile apps for observability, we knew we had to pull the data and drive some points home. Many versions, device fragmentation, no hot fixes or quick rollbacks for releases, delayed data…oh my. But don’t be scared.
To better understand and communicate this complexity, we analyzed anonymized data from our platform — sourced from apps using our SDK and observability platform — and compiled it into a mini-report: “Mobile Complexity in 6 Charts.” Here’s a quick look at what we found:
- Apps are juggling dozens of versions. A quarter of apps we observed had more than 10 active versions in production — and 6% had over 50. Supporting that many codebases adds tremendous overhead.
- Devices are wildly fragmented. Android apps in our dataset saw usage across thousands of unique device types, while iOS apps weren’t far behind. Each hardware combo brings performance unpredictability.
- Network variability is the norm. User connectivity varies dramatically by platform and geography. For instance, in Africa, half of user sessions were on cellular, compared to only 25% in North America.
- Monitoring is delayed. It’s not just complexity — it’s visibility. 17% of telemetry data arrived over an hour after a session ended, and 7% took over 12 hours, delaying incident response and insights.
These numbers aren’t just stats — they’re a wake-up call to the entire engineering organization. The best teams are building bridges and fostering a deeper understanding of modern observability, shedding light on the unique needs of backend and mobile teams to better serve their end-users. To build stable, performant mobile apps, you need real-time visibility into real-world conditions. That’s where Embrace comes in. With a common language enabled by OpenTelemetry, semantic conventions to use across the stack, and expertise in tooling that reaches far beyond traditional solutions, we’re here to enable the art and science of app performance and real user monitoring for your organization.
? Read the full report
? Get a demo to see how we help teams cut through the noise.
