How Does Your App Compare to the Top-Tier?

In the age of social distancing, mobile apps have provided consumers a variety of ways to carry out daily tasks with little to no contact. As a result, there has been a significant boom in popularity for mobile application usage, especially in the categories of food delivery, restaurants, and grocery.

We recently examined how well these mobile apps are meeting consumers’ growing expectations and demand for optimal speed and performance. Our research ranked the performance of 25 shopping, restaurant and food delivery apps in five major COVID-19-impacted cities, across popular iOS and Android devices, on major wireless service provider networks. Some of the applications we investigated include, Amazon, Walmart, Burger King, Grubhub and UberEats.

Is my app delivering optimal digital experiences?

The first step is understanding what key performance indicators (KPIs) amplify a user’s experience. The next is how to optimize these KPIs to improve conversions, reduce churn, and ultimately achieve faster time to market. Our benchmark is based on four KPIs for contactless e-commerce apps: load product time, add to cart time, launch time, and search time.

 Our observations include:

Launch time was painfully slow for all but the speediest apps - The study found that the best contactless e-commerce apps took under two seconds to load. However, the average launch time for all apps was 4.1 seconds—a poor overall score that points to a need for launch time improvements. A few contributors to poor launch time include launch time impacted by multiple TCP and TLS connections, multiple HTTP errors, or high battery drainage.

Newer apps, like Shop, excel at Search - Consumers more often than not have something in mind to buy when they begin their e-commerce app experience. Therefore, the search function is crucial to any application. Surprisingly, some of the largest retailers—such as Amazon—featured slow search times (negatively impacting the average), while the relatively new Shop app from Shopify excelled across all metrics, performing 5.5x faster than Amazon in returning search results.

What are the top contributors to app latency?

Now let's take a look at a few major contributing factors to an app’s slow performance.

• Due to slow TLS with lots of connections to multiple Amazon hosts, the Amazon app took twice as long to launch compared to other top apps including, Home Depot, Kohl’s, and Best Buy. One recommendation we would have for Amazon is to consider enabling keep-alive to reuse established connections.
• Duplicate requests can lead to prolonged launch times, as was the case with the Postmates app, which took four times as long to launch on iOS compared to Uber Eats. Postmates should consider caching their HTTP requests.
• The UberEats app gets to the core user experience the fastest, while others such as Grubhub, took four times as long to load on Android. This was mainly attributed to multiple calls to 3rd party hosts for initializing SDKs. We recommend reusing connections to avoid the overhead of establishing a new one and examining if the number of SDKs integrated into the app can be reduced or loaded later on to improve launch time.
• The Pizza Hut app experienced lag times on Android for two reasons: the slow performance of its backend server and large image files. To address this, Pizza Hut should look into the amount of server-side work being performed before requests are fulfilled. Especially when loading images, using a file like JPEG or WebP instead of PNG will reduce the file size and help load images faster.

There are several ways that businesses can ensure their apps deliver optimal performance. By intelligently automating functional, performance, and load testing end-to-end across native, hybrid, and web applications, on real devices and real networks before, during and after launch, organizations can detect build-over-build regressions early on, and be alerted to high priority issues. By monitoring and baselining live app KPIs by location, device, OS and carrier networks organizations can identify areas in need of improvement, as well as analyze unique performance and UX data to enhance their apps’ digital user experience.

We’ve always relied on our devices, but social distancing has made that reliance more of a dependency. Despite social-distancing restrictions lifting across the country, and around the world, consumers will continue to use these apps to mitigate health risks and continue contactless practices. As organizations small and large continue to shift to web and mobile applications, those who fail to deliver on speed and performance will lose in the end.