Mobile testing is hard. But with today’s application architectures, load testing can address many of the important bottlenecks in your overall app or browser performance. First, though, let’s look at user behavior.
In Comscore’s latest report, the company shared trends coming into Q4 of 2017 - and shared a few insights about consumer behavior that are relevant to our work making the fastest, most usable apps and sites possible.
It turns out that while e-commerce growth remains strong (more on that in anotherarticle), mobile commerce, or m-commerce, remains steady as a percentage of that overall e-commerce growth. In other words, while it might seem intuitive that mobile commerce would grow more quickly, it remains at about 23% of all e-commerce. Thus, m-commerce is growing, but its growth is because of the overall growth of e-commerce, not because individuals buy more things via their mobile devices at the expense of another device (laptop, PC).
What that percentage doesn’t tell us, though, is how your customers behave before that ultimate purchase. They may be three times more likely to make an e-commerce purchase using a laptop or PC than using a smartphone. But they’re very likely using your site to research and consume information about your products before they make that ultimate purchase.
Those interesting data points tell us this: regardless of the front-end client your customer uses, you had better be testing the whole system. Your web-based or app-based front ends are important, but the entire app architecture is involved, including server-based, cloud-based, data-related and API access code.
That’s where load testing comes in. Load testing allows you to test that speed across front ends, uniting the app architecture and simulating user experience. It’s important to test mobile client responsiveness, for sure. But testing your web-based client regularly and consistently, with significant demands on your whole architecture, will yield the speed results you need for rewarding e-commerce and m-commerce.
Plus, based on these insights, you can create more comprehensive testing scenarios. For example, if your e-commerce usage patterns match those Comscore reports (about 3:1 “desktop”:mobile), yet you’re receiving more mobile-based traffic than those purchase results would suggest, you might consider emphasizing non-checkout-related tasks in your mobile experience. Instead, you might emphasize performance around product specs. Or you might test for “shopping list” addition speed, or ease of login for “save for later” queues. If research and content reflects mobile usage, image and cache optimization - as measured with load testing - may be paramount.
No matter how you set up your load tests, be sure to see if your usage patterns might match those Comscore reports. If so, tailor your tests and optimize your results accordingly.