Top 4 Myths of Website Performance Testing

performance testing myths
When it comes to the success of a website, performance testing at realistic scale is key.

Without exercising web and mobile apps and infrastructure with the same activity and traffic levels that you expect from real users, how can architects, developers, testers and operations personnel prove it will be an effective, profitable resource?

To be completely frank, they can’t. Performance testing is an absolutely critical part of rolling out a new website, and testing at full scale – whatever that means for your site – is imperative.. However, there are still a number of myths circulating throughout the industry that hamper effective testing efforts. Let’s take a look at these misconceptions, and explain from an expert perspective just what makes them erroneous and misleading.

Myth #1: The website is good as is – it doesn’t need performance testing

Simply put, this belief is absolutely false no matter how successful a website seems to be. Forward-thinking organizations change the content and capabilities of their online platforms on a consistent basis, adding new products and different ways for customers to connect with the business. When these changes take effect, it is essential to make sure that they work correctly and perform to developers’ expectations.

Even if decision-makers are convinced of the website’s success, testing should still not be taken out of the equation. After all, when testing falls short, and sites fail under load, it’s these decision makers who face the heat, and will be seeking answers! The internet is full of stories of companies and organizations who thought their site was ready, but clearly was not. No one will forget the woes of Healthcare.gov!

Myth #2: Performance testing is far too expensive and won’t fit into the project budget

While it may have been true in the past, the cost of a performance testing solution is no longer prohibitive today. Although previous testing strategies required the use of expensive, user-constrained software and elite coders to make use of these resources, this is not the case anymore.

Cloud Testing Fundamentals We at SOASTA provide affordable website performance testing that delivers all the capabilities required for enterprise Web and mobile load testing in packages that make everyday, full-scale testing affordable. On top of that, the technology is highly usable, allowing new testers to be effective quickly while still enabling the most experienced Performance Engineers with everything needed for complex end-to-end production testing.

Myth #3: Performance testing can only take place at the end of the project.

Another completely misleading – and potentially harmful – falsehood is that performance testing should only happen at the end of the project, when it’s “ready”. Indeed, this is how load testing has been done in many traditional shops and typically because of the time and cost associated with load testing. But, this approach can lead to a number of challenges, including the need to go back and almost start all over if testing shows that the website’s architecture is faulty. Instead of wasting valuable time and resources by testing at the end, organizations should leverage a performance testing solution throughout the process to make correcting any issues or oversights much easier – and less costly. To get ahead of the increasing cost of performance bugs found later in the development cycle, more and more companies are adopting a continuous performance testing approach, even implementing small performance tests during early stage development.

Myth #4: I can test my website with open source or my own custom code

Some teams think that they can hammer out a site accurately with custom HTTP calls or an open source solution. While there are certainly ways to test with costless tools, “free” is a relative term. Whether open source or custom-coded, the problem here lies with the fact that the emphasis and most of your valuable time is spent hand coding the tests. Valuable engineers are tied up writing test code instead of actually executing tests and fixing bugs. More troubling, most free approaches completely avoid the most important part of testing – results. That is due to the fact that collecting and reporting on load testing results is difficult. SOASTA is dedicated to eliminating unnecessary coding and other non-productive testing tasks, like administering load generators or de-bugging test code. A fast and effective visual approach to test building that accurately captures the paths real users take in your system, coupled with our Global Test Cloud that provides access to thousands of worldwide cloud servers, takes most of the dev and setup time out of load testing. Most importantly, ours is a big data platform focused on the real-time analysis of web and mobile performance. This means that not only is the test development time compressed, but results are presented during tests, often eliminating the dreaded re-test. When issues are found and fixed during test run-time, your site will be up and ready for anything, faster.


This article was originally published on SOASTA’s Blog and was reposted here with permission from Brad Johnson.


About the Author:
Brad JohnsonBrad Johnson – VP Product Marketing, SOASTA – Brad is a cloud-testing pioneer who joined SOASTA in December 2008. His former roles as head of test and monitoring products at Compuware, Mercury Interactive and Borland prepared him well to disrupt the skeptical and established software quality market with updated approaches and technologies for continuous web and mobile testing.

You have two opportunities to come see Brad at STPCon, he will present his session entitled: “Jane, Stop this Crazy Thing!”. Brad is also one of our Keynote Speakers this Spring – The Measures of Quality. Join us in San Diego!