From Anaheim to the Amazon, Monitoring End User Experience of Websites That Rely on a Public Cloud

New research from comScore, indicates that globally, one in every five internet users visited Amazon’s various websites during June – congratulations Amazon, that’s pretty cool. The study points out that Amazon has an impressive reach with no single geography contributing a lion’s share of visitors (compared to Walmart.com, for example, with over 80% of visitors coming from North America.)

 

Worded as “Amazon various websites” makes me think these statistics must include traffic generated by Amazon hosted services. And if memory serves me correctly, Amazon Web Services is the largest public cloud provider out there. This leaves me to wonder, how do public cloud providers ensure their service results in a high quality user experience.

 

As more and more companies in geographies across the globe rely on a hosted services model, it’s important to consider how you will gain and maintain independent visibility into the performance of the service you rely on. This is critical because the service provider (Amazon or otherwise) will have a natural bias towards not highlighting performance problems that negatively impact user experience. So you’re left to wait until your customers bring this to your attention, meaning hours of customer service time attributed to soothing frustrated users, costing you time and money.

 

So the ideal solution would be a proactive one, right? The ability to detect issues early, and determine the characteristics they share, such as geographical source, browser type and mobile device, can be the difference between a minor issue and major business impacting outage.

 

One way to stay ahead of business impacting performance issues is to passively monitor the real end user experience on your hosted web applications. (Shameless plug here – Foglight End User Monitoring Solutions can do this.) One of the most valuable perspectives for many organizations is a geography aware dashboard that provides an at-a-glance view of application health on a map and incorporates alerting to call out any performance problems which may be impacting user or customer experience. This perspective can also help you isolate the geography or other factors of commonality where users are experiencing poor performance, without relying on users to report issues or having a synthetic robot preconfigured to look for that particular issue. Using this capability, a large financial institution was able to identify that users in Seattle were experiencing slow page load times and attribute the poor performance to their cloud provider. Having this information at their fingertips empowered them to have a more productive troubleshooting conversation when they called to report the issue.

 

Amazon is leading the world in users visiting their sites. If your web application and business relies on Amazon or any other public cloud provider, you need to consider how your end customer experiences the service, how you can measure it, and how that impacts your business.

About the Author