When it comes to web site performance, retailers must guard against having tunnel vision. All too often site performance is measured simply as speed of page downloads and site uptime, when in reality there are a slew of other key performance indicators that retailers often overlook or may not even be aware of.
A recent survey by Harris Interactive conducted on behalf of Tealeaf Technology Inc. revealed that 42% of respondents that experienced a performance problem when shopping went to a competing retailer’s site or abandoned the transaction.
Performance problems other than speed of page downloads that can drive shoppers from a retailer’s site include links to an incorrect page or product, inability to add an item to the shopping cart, delays during checkout, incorrect page feeds, and poor navigation paths, all of which can negatively impact the shopping experience and cause shoppers to abandon the site and shop on a competitor’s site.
“Most retailers look at performance in a vacuum,” says Geoff Galat, vice president of marketing and product strategy for Tealeaf, provider of online customer experience management applications. “They don’t realize there are thousands of other little performance problems that can be related to site design, programming and search that negatively impact conversion.”
Creating the balance
Identifying the little problems that harm site performance requires having a baseline for performance, something many retailers neglect to establish. Consequently, their view of site performance is subjective, which can lead them to focus on the wrong issues, as opposed to weighing how each problem they identify can impact their business.
“Retailers need to view performance monitoring strategically rather than tactically,” says Matthew Poepsel, vice president, performance strategies for Gomez Inc., provider of web site performance monitoring services. “When performance is viewed tactically, it can lead to blind spots and over investment to correct problems that have little impact on the shopping experience or profitability. When performance is viewed strategically, retailers can align performance metrics to their business goals and focus on the metrics that help them achieve those goals.”
Lining up priorities
Aligning performance metrics with business goals starts with prioritizing the importance of each measurement. “Retailers need to ask themselves how each performance metric impacts conversions, abandonment rates and sales,” says Pedro Santos, senior product marketing manager for Akamai Technologies Inc., which has a computing network to accelerate retail web sites and applications.
Instead, many retailers put their energy behind fixing programming code or infrastructure problems, such as their connections to Internet service providers, even though that may not be the culprit, according to Santos. The reason is that in-house IT staffs are oriented to writing programming code and integrating the hardware infrastructure so that is where they focus when it comes to addressing performance issues.
As a result, retailers get bogged down adding infrastructure rather than looking at the Internet “middle mile” for performance, an area that has the biggest impact on customer satisfaction. “Retailers will spend a lot of money on fixing programming code and hardware infrastructure and still end up with poor site performance,” says Santos. “Until they look at all the performance problems areas and how they impact conversions and sales, it is difficult to figure out what to fix.”
Prioritizing performance metrics requires good communication between customer service, IT, marketing and merchandising about performance problems brought to their attention by customers or that are detected through site and load testing. “Successful retailers communicate performance metrics across all departments and with their technology partners,” says Poepsel. “Communication makes it possible to not only prioritize performance metrics, but also understand what the data says and to identify areas where improvements can be made and what kinds of dollars those improvements can yield.”
Once a performance problem is identified, retailers need to look at what the customer experienced when encountering the problem. This granular view helps retailers determine how widespread the problem is across the customer base. “There are problems that affect only a small percentage of customers. Without the ability to go back and review the customer’s shopping session, identifying the problem and the extent of it is like looking for a needle in a haystack,” says Tealeaf’s Galat. “Retailers have to be able to see what the shopper experienced that prevented them from converting or prompted them to abandon the site before they can properly assess the extent of the problem and correct it.”
Tealeaf’s survey on how shoppers respond to performance problems revealed that 31% left a retailer’s site and shopped at a competing site after encountering difficulty navigating the site. 30% said they experienced endless page loops that prevented the completion of a transaction. 14% said they abandoned the shopping session after receiving an error message.
Bad track record
One of the most surprising results to emerge from the survey is that 52% of respondents that contacted a customer service agent after experiencing a performance problem said they did not get their problem resolved and completely stopped doing business with the retailer.
“Getting a user session to the service agent can help them identify and resolve problems faster, because they can see the path that led to the problem,” Galat says. “The aim is to get beyond the quantitative to the qualitative to find out why a problem occurred and how many shoppers it affects when troubleshooting a problem. Once retailers know this data, they can determine the cost of the problem to their business, which helps them prioritize.”
The growing sophistication in how retailers identify and prioritize performance problems applies to speed of page downloads. As more retailers embrace rich media applications, such as Ajax, page weights become heavier, which slows the rate at which they are fed to a shopper. Specifically, web sites designed using Ajax require a higher than normal volume of data requests between the front-end web browser that delivers pages to the shopper and the back-end server that stores the page content, according to Akamai’s Santos.