The average return on Facebook ad spend rises 26% in Q3, according to social media advertising firm Nanigans.
(Page 2 of 2)
"There are a lot of feature phones still in use that have limited browser capabilities compared to a smartphone," Hansen says. "It's important to map the user interface and performance of the web site to the capability of the mobile device so the site meets consumers' expectations. The frantic pace of innovation in the mobile world is good for retailers because it means they need to be paying attention to the space."
Testing, for real
Real-time testing is considered the best way to accurately gauge the performance of a mobile site. While many retailers perform simulated load testing, the process does not recreate all the speed bumps shoppers are likely to encounter in the mobile world.
"There are a lot of variables that are not always taken into account with simulated testing, such as whether the consumer is using the latest version of their operating system's browser, which means no simulated test script can cover all the potential variances," Moran explains. "Similar variances exist for consumers accessing a web site from a desktop or laptop computer. The only way to discover how these variances affect performance is to track the actual performance at the end user level."
New Relic conducts real-time, end-user application and site availability monitoring through a software-as-service or SaaS-based solution, which means New Relic hosts the application and clients access it via the web. Retailers can use New Relic's tools to gain visibility into, and measure the performance of, their applications developed with such programming languages as PHP, Ruby, .NET, and Java.
Other tactical tips for enhancing site performance include knowing how to manipulate image size and quality. While large, high-resolution images pop off the screen, they are dense pieces of content that can significantly slow page downloads. Testing several sizes of the image and measuring how consumers respond to them allows retailers to zero in on the image's optimal size to ensure speedy download times.
Image resolution can also be tested to determine how it affects customer behavior. Lightening background colors or thinning the background detail on an image of someone holding a book toward the sky, for example, can speed the download time for the image without noticeably affecting its quality.
"Image size and quality need to be tested to determine the right balance between faster downloads and the impact on customer behavior," SiteSpect's Hansen says.
For all the strategies retailers can employ to improve web performance, one potential trap to avoid is over-investing in high-speed servers or faster databases. "Adding more hardware is considered an easy fix, but it is costly because it does not address the issue of optimizing the code for the web site," New Relic's Moran says. "The real investment in site performance is in the commitment to end-to-end optimization."
A key part of making a strong commitment to site performance includes designating someone in the organization to be directly responsible for performance.
"Someone needs to own the process of continually optimizing site performance because every opportunity to improve the user experience may come at the expense of speed, and that's a real risk," Hansen says. "Performance tools will only get a retailer so far. It takes a dedicated individual to interpret performance data and act on it."