With 8 million customers around the globe ordering business cards, letterhead and other printed materials from its online venue, VistaPrint executives understand the power of Web analytics for uncovering customer trends. Getting the most of out their data was a matter of measuring -- and acting on -- the right metrics.
It took Nick Ruotolo, VistaPrint's vice president of analytics, several months to realize that general site metrics, such as the conversion rate, drop out rates, and average order value, weren't truly actionable. These metrics were merely outcomes, not the means to the outcome. Ruotolo now has a new analytics strategy: measuring the performance of very specific critical site paths.
"We use conversion funnels to track users' step by step success rates as they traverse these paths," he explains. Conversion funnels demonstrate each step of the conversion process, and pinpoint where visitors dropped out of the conversion process. By applying this technique, Ruotolo says he has identified very specific site improvements that have made positive impacts on VistaPrint's bottom line.
Keeping an Eye on the Prize
Focusing on key, actionable metrics led VistaPrint to test a redesigned version of its file upload path that aimed to streamline the process. When the conversion rate didn't change, Ruotolo didn't stop testing. Instead, he continued drilling down into the metrics. He hit the analytics jackpot when he compared the dropout rates of each step of the navigational paths and discovered that the redesigned version had a much higher dropout rate on one of the steps.
"We were then able to focus our attention and subsequently fix that one page of the path, resulting in a significantly higher conversion rate -- about 5 percent -- for the redesigned version," Ruotolo says. "Had we only relied on the high-level conversion rate as a metric, we never would have realized that opportunity."
VistaPrint's story illustrates a critical point in Web analytics: Enterprises could be leaving money on the table -- or losing out on opportunities -- by focusing on the wrong metrics. Of course, there is no formula, per se. Deciding which metrics to act on can vary from site to site and from industry to industry. However, there is a rhyme and reason to choosing metrics. The key is understanding your organization's unique goals.
Some Baseline Metrics
If you are just getting your feet wet with Web analytics, there are some industry-specific metrics you could start with. For example, if you are trying to generate sales leads, then metrics such as leads generated, lead conversion ratios, cost per lead, number of Web inquiries, and Web inquiry failure rate would be important to track. By contrast, if you are running a financial services firm and hope visitors will submit electronic application forms, then you'd want to measure metrics like completed online applications, application conversion rate, number of self-service transactions, self-service failure rate, and self-service penetration.
Online media and content sites tend to look at altogether different metrics. These reader-oriented sites would record pageviews, monthly unique visitors, pageviews per visitor, visits per visitor, and number of subscriptions in attempts to impress potential advertisers. Travel and hospitality sites look at a completely different set of numbers, such as revenue per booking, look-to-book rations, sales cycle rate, origin and destination city pairs, and campaign affinity. Finally, e-commerce sites' essential measurements would fall in the realm of total revenue generated, average revenue per visit, number of orders, average order value, and order conversion rate.
As you can see, there are many different metrics. These are just a few of scores. Whether you use these or some other metrics, the key is to determine which statistics offer the deepest insights into the performance of your site, then keep an eye on those stats at regular intervals. That may be daily for large enterprises running multiple campaigns or monthly for lead generators with static offerings.
What you don't want to do is get overwhelmed with numbers that make your site look popular on the reports, but might not pay off on the conversion end. "Too many sites are focused on the popularity contest," says Chris Parkin, director of Product and Solutions Marketing at Omniture, an online business optimization software and services vendor based in Orem, Utah. "If you focus on optimizing the five performance metrics for your industry," Parkin says, "you'll recognize positive business gains."
What Really Matters?
Organizations often report on Web data like page views and clicks simply because the data exists. This can be limiting, says Richard Foley, who is the worldwide product manager for Web Analytics at SAS, the $1.9 billion business-intelligence software provider based in Cary, N.C.
Foley says people tend to focus on metrics like bytes, page views and clicks rather than on what really matters: the customer experience on the site and getting customers to complete a specific business objective you've defined.
"In the mid-90s, some companies reported hits and page views as a way to show their worth to investors while not stating how many unique visitors they had," he recalls. "Today, many companies are looking at Web 2.0 activity and doing the same thing: focusing on the wrong metrics and measures."
Foley recommends doing just what VistaPrint did -- focus on conversion funnels. The idea, he says, is to persuade visitors to take a desired action on your Web site by offering them a rich user experience. That rich experience may include design, product affinities, content, robust search or some other factor. In Foley's view, measuring this process through conversion funnels is key to success.
Leaks in the Funnel
With the conversion ideal in mind, once-popular metrics, like length of time on site may not always be the most relevant statistic to measure. In many cases, it's not how long a visitor is on the site that matters. It's what they do on the site that counts.
"A visitor might spend a long time on your site not because they are a good customer or love your offerings. Rather, they might not be able to find what they need quickly," Foley explains. In the latter case, you may be looking at metrics like duration and pageviews as positives when in fact the customer is frustrated by your site and may not return.
So how do you take action on your meaningful metrics? VistaPrint's Ruotolo paved the way with his redesign test and optimization process. And, Omniture's Parkin offers up related advice for retail industries. If the average revenue per visit as a metric is trending down, he notes, it could indicate a conversion problem on your site. Your job is to find out why.
"You should look for leaks in your conversion funnel," Parkin says. "If your conversion funnel is converting well, then it would be good to check the pricing of your products relative to the competition."
Indeed, there are many factors that can affect a Web site's performance. Monitoring Web analytics metrics can be like picking up clues in a mystery. One clue may tip you off to part of the solution, but it's likely that you'll need to test your hypothesis to prove your case. Ultimately you'll find that measuring the right metrics -- metrics that are in line with your business objectives -- can shorten the investigation and ultimately put more profits in your pocket.