The Truth About Network Performance: It's the Applications, Stupid

Posted by Jim Rapoza
September 05, 2012

A common phrase heard during election seasons is, "It's the economy, stupid"--meaning that, while other issues may make headlines at different times, in the end, elections are about the state of the economy. And when it comes to network performance, IT administrators and the vendors who support them are realizing that it's an equally important truism when it comes to applications.

Traditionally, tools and products focused on network performance have dealt mainly with metrics. Things like bandwidth availability and packet loss tended to dominate the discussion about network performance. And the products that vendors promoted used similar verbiage about "boosting network capacity" and "optimizing WAN performance."

These are certainly core elements when it comes to providing a reliable and high-performance IT infrastructure. However--and I hate to break it to you, network guys--but outside of your data center, no one else in the company really cares about network performance. Even if Mary in sales complains that the network is slow, what she really means is the application she's using is performing poorly.

If you think about it for even a second, this makes perfect sense. The main reason that networks exist is for the applications that run over them. And the dynamic changes occurring in businesses today make having a focus on applications even more important.

The increased use of cloud (public, private and hybrid) is changing how applications are delivered and used within enterprises. And virtualized application infrastructures are making it vital to focus on applications rather than servers.

Also, since many different types of applications come in over Internet protocols, the old approach of optimization based on protocol type doesn't make sense. For example, businesses need to ensure that their Web-based CRM apps perform well while also ensuring that Facebook usage doesn't eat up bandwidth, even though both are coming in on the same network port.

The increased focus on the importance of applications should help boost the strategies and technologies used to improve application performance and reliability, and provide an improved end-user experience. And we are definitely seeing an increased focus.

At Interop 2012 in Las Vegas this past spring, nearly every vendor there was taking an application-first approach, even those mainly focused on traditional network hardware. And a quick look through the many marketing and press releases flowing out of network vendors these days will often display a high focus on applications and providing a good user experience.

But this is more than just marketing-speak, and it means a lot more than just how vendors position themselves. End-user businesses that focus on application performance and improving user experience and satisfaction over traditional network performance metrics often see an end-to-end boost in all performance areas, including high levels of application adoption, uptime and user satisfaction with application experience.

In my research, I've found that leading businesses that have the highest levels of achievement in overall performance issues also have a high focus on application and user experience. According to the report "Application Performance in Complex and Hybrid Environments," businesses that achieved in the top 20% of all surveyed companies were more focused on strategies such as improving application usability based on end-user experience and on tracking performance of critical applications.

To a certain degree, this can seem like a pretty basic and obvious thing. But for organizations that need to ensure good performance for applications deployed to their employees and customers, it's actually a pretty significant change in perception.

And it is a welcome one. When companies focus on applications and user experience, it benefits not only the applications but also the networks and data centers that they run on.

Related Reading