The following is a guest post by Ozan Unlu, CEO of Edge Delta.
If you’re a fintech offering personal finance, digital banking, or investment trading mobile app, you know the importance of high performance (speed and reliability).
You must also know that the bar for user satisfaction with mobile app performance is increasing daily.
Studies have shown that humans are most comfortable, efficient, and productive with less than two seconds of electronic response times since waiting more than two seconds forces users to concentrate up to 50 percent more to retain their attention.
RELATED: Deep Dive into the New BBVA Mobile App
As a result, 48 percent of users will uninstall an app in response to slow performance. Moreover, the higher the percentage of users experiencing crashes, the higher the drop in app store ratings. A two-star rating drop typically results in a 50 percent lower user acquisition rate.
In this context, superior mobile app performance is a must, especially for users of fintech apps, since these are being used for sensitive and urgent purposes related to managing and handling users’ money.
Observability is key
Observability measures a system’s current performance in IT based on its generated data. It is critical to achieving end-to-end monitoring and management of mobile app performance. This includes detecting issues the organization may not have anticipated, which is essential given the increasing reliance on dynamic cloud environments where just about anything can happen.
Log data plays a huge role in observability initiatives. A log is a computer-generated file that contains information about usage patterns, activities, and operations within an application.
Understanding the overall app behavior and everything happening inside the app is essential. Fintech apps feature numerous database calls to deliver users’ current account balances and additional API calls to process payments, execute trades, and more.
These activities generate tons of log data.
As log data is produced at such an incredible rate, organizations find it harder to get their arms around it to identify anomalies and growing hotspots, which can sprout up virtually anywhere.
Established banks often have the luxury of being able to afford enterprise observability platforms, where they ingest all this data for searching, monitoring, and analysis.
Emerging fintechs, however, may not have access to such platforms, and even when they do, the ongoing viability of this “centralize and analyze,” “store and explore” approach is highly questionable, given:
- The costs required to ingest and store exploding data volumes – The threat of exceeding a storage limit – and getting unpleasantly surprised with a massive bill as a result – forces emerging fintechs to make painful, often indiscriminate decisions of which data to keep and which to neglect, leaving them with significant blindspots that increase the chance of an unplanned downtime nightmare.
- Undue developer stress – To avoid any further consequences associated with delaying troubleshooting, this task has to be tackled immediately, regardless of the time of day. If data is incomplete (i.e., datasets have been neglected), it becomes even more stressful, and it’s no wonder that the stress involved in troubleshooting is one of the leading causes of mobile app developer stress.
- Latency in analysis – The time needed to centralize data can inhibit real-time query speed. Furthermore, centralization results in too much data for your app developers to sift through as they proactively monitor and troubleshoot the app’s performance. When it’s 6 a.m., and your help desk gets angry calls from users that your app is crashing, time wasted sifting through data will not be acceptable.
A mobile app’s observability data is pure gold for saving time (and face) in the long run by revealing bugs and other issues with fintech apps. But with the aforementioned challenges of the traditional “centralize and analyze” approach, what changes may be in order?
- Focus on the Source: This means flipping the “centralize and analyze” approach on its head and permitting data to be analyzed and processed as its source, at the edge, as it’s being created. This helps avoid excessive storage costs and analysis latency due to centralizing while automatically surfacing anomalies and relevant loglines, so developers know the root cause of the issue instantaneously.
- Move from a “Big Data” approach to “Small Data”: When observability is decentralized, datasets can be processed in smaller amounts, in real-time, and in parallel. This ensures there’s an eye on all data and no datasets are neglected while preserving real-time query speed.
- Make All Developer Data Accessible: It may be impossible to say an engineer will never be woken up in the middle of the night, but if this happens, they will have access to all the data they need and achieve much faster root-cause analysis, thus minimizing their stress and burnout. By decentralizing accessibility, the notion of data limits becomes irrelevant, and this should also apply to pre-production data, which offers a wealth of insights that developers can use to stave off production problems in the first place.
Conclusion
For virtually every traditional financial service, a fintech company innovates to provide an improved user experience – including exceedingly fast and reliable user interaction.
The challenge is that as the internet grows more complex, relying more on the cloud and serving more users in more geographies at the edge, delivering such slick and polished experiences becomes much harder.
The potential issues that can cause an app to crash in this modern, cloud-based world of interconnected systems and robust applications are too numerous to count. This is especially true for emerging fintech apps.
In this context, observability needs to be implemented in new ways.
By decentralizing and pushing analytics upstream, closer to the data source, fintechs are in a stronger position to harness all their observability data in a much more efficient way and put that data to work for them in the ongoing effort to maintain and support the performance of their mission-critical applications.