As part of a plan to cut costs by $500 million, Staples says it plans to close up to 225 North American stores by ...
SPONSORED SUPPLEMENT: Web Site Analytics: Measuring the Little Things Can Mean A Lot
A detailed look at how customers use a web site can yield profitable changes
When executives of Outrigger Enterprises Inc., operator of hotels and resorts, wanted to learn how customers were using Outrigger.com, they analyzed page load times, abandonment rates, how much effort it took a visitor to arrive at a web destination and how customers moved through the site.
In the analysis they noticed that many would-be travelers clicked on the Reservations button, entered information, viewed results, then clicked on the Back button and went through the entire process over again. What, they wondered, were customers doing?
Further analysis showed that customers were trying to find room prices. Outrigger.com did not show prices of rooms until a customer made a reservation. So customers would make a reservation, find the price, then, if they were looking for something more suited to their budget, they would start over again to find a different price. “The reservation engine was being used as a search engine,” says Bill Sthay, vice president of interactive commerce at Outrigger.
Called “pogo-sticking” because customers jump back and forth between pages, such creative use of web sites by consumers is common. Such subtleties of web use, however, are not always apparent to web site operators, be they retailers or hotel operators. But now as merchants’ understanding of the web grows more sophisticated, so do the tools to analyze the usefulness of the web.
Once, the measures were number of visitors, conversion rates and percent of shopping carts that customers abandoned. Now the measures are more finely tuned and focus on such metrics as how many shoppers left at a certain point, or how did shoppers arrive at a certain page, where did they go from there and what did they do. “When you look at measuring success of a web site, you need to ask what you want people to do at your site that will improve results for your business,” says Jim Plymale, vice president of marketing for WebCriteria Inc., a Portland, Ore.-based provider of systems to measure customer behavior on web sites, then guidance as to how to improve the customer experience. “If I’m a retailer, I want someone to buy something, so I want to know how I move someone closer to accomplishing that job.”
WebCriteria is one of a new breed of vendors that provide detailed measurements of web site metrics, known as clickstream analysis. Analysis tools receive a steady stream of information about how customers are clicking around a site and what they’re doing. Based on the parameters that the retailer sets, they use the data to determine where there are problems at the site, then help the retailer decide how to fix the problems.
For instance, in analyzing clickstream data at Camping World Inc.’s CampingWorld.com, retailer of RV accessories, managers found customers abandoning the shopping cart almost at the end of the checkout process. Because that was the point at which the site displayed shipping costs, they reasoned that customers suffered sticker shock and departed. Now CampingWorld.com displays the shipping costs upfront in checkout. That change has resulted in a $900,000 boost in annual sales, Plymale says.
Pressured by ROI
As web sites focus more on ROI, the web analytics market will benefit, reaching $1 billion by 2006 across all industries, according to Jupiter Research. Jupiter reports that 48% of retail sites employ some sort of analytics today and that number will reach 58% in 2006. The market will be powered by just the sort of experience that Outrigger.com had. “As managers are increasingly pressured to provide insight and optimize the value of their web properties, they will be forced to adopt more flexible tool sets and sounder analytical methodologies,” says Matthew Berk, analyst at Jupiter Research.
As useful as clickstream analysis is, some analysts argue that it’s not enough. Retailers also need insight into the shopping process that shoppers themselves provide in order to make strategic decisions about web site changes. “Clickstream analysis is important but it doesn’t tell me if my customers accomplished what they set out to accomplish; whether my site met their needs and exceeded their expectations,” says Larry Freed, president and CEO of Ann Arbor, MI-based ForeSee Results Inc., a company that uses the methodology behind the American Customer Satisfaction Index to quantify customers’ satisfaction with web sites and correlate it to their likelihood to buy from the site again. “One person may click on four pages and another on 27 pages, but that doesn’t tell me which is the better customer. The person who clicked on only four pages may have found what he wanted right away, then gone offline to buy it.”
The American Customer Satisfaction Index has proven that a direct correlation exists between customer satisfaction and success of a company. ForeSee Results interviews a fraction of customers at a site, then applies formulas to measure satisfaction, sources of dissatisfaction and future likelihood to buy. “Just looking at behavior doesn’t predict the future,” Freed says. “You have a difficult time telling where your biggest payback will be.”
The American Customer Satisfaction Index was developed in 1994 as a cross-industry measure of satisfaction with the quality of goods and services in the United States. It seeks to link customer satisfaction to future consumer behavior and economic returns. It applies a mathematical model that eliminates anecdotal reports of satisfaction, customers’ self-reported levels of satisfaction and customers’ assessments of what’s important in making a buying decision. Rather, it asks a series of specific questions from which analysts derive overall satisfaction levels to determine which changes will have the biggest impact.
So in the case of a retailing web site, rather than ask a customer if he was satisfied with the ordering process, ForeSee will ask a series of questions about steps in the ordering process, such as: Was the ordering mechanism easy to use? Were you satisfied with the number of shipping options? Was the payment process easy and were there enough options? Was the order confirmation adequate? Was security acceptable? From the answers, ForeSee derives a customer satisfaction measurement and from that computes customers’ likelihood to return as well as their likelihood to be offline customers.