E-commerce Link: Taming Web Analytics
Most Web analytics programs churn out a mountain of data—some of it relevant and much of it not. The challenge marketers face is how to sift through that mountain to uncover the nuggets of information that will help guide e-commerce design and decisions, and ultimately drive profits.
When Forrester Research recently polled members of its Web Analytics Peer Research Panel, 53 percent responded that taking action on analytics findings was the most significant nontechnical hurdle they faced. At the same time, Forrester says analytics was the top spending category for research professionals in 2006. That’s because there is a proven business case for investing in Web analytics. According to Forrester, analytics can improve the ROI from redesign projects by 10 percent to 25 percent by identifying problems with site design, such as drop-off at checkout. The same report shows analytics can improve the ROI of marketing efforts by 10 percent to 30 percent by identifying campaigns that are underperforming and highlighting those campaigns that yield higher results.
So how do you map analytic metrics to strategies that drive revenue? Let’s consider the conventional data marketers usually examine to assess Web site performance, and then explore some overlooked findings that might point to a concrete plan of action that yields real returns.
What’s Wrong With Conversion and Abandonment Rates?
Traditionally, retailers have used two basic metrics to gauge success: conversion rate and shopping cart abandonment. While these provide a good macroview of online retail performance, they only provide a speedometer that tells you how fast you’re going. To give online merchants a look under the hood, MarketLive recently compiled the first MarketLive Performance Index, collecting key metrics from leading e-commerce providers to generate benchmarks for a range of analytics, and highlight which metrics are worth tracking for real insight.
For example, Forrester research indicates that the average conversion rate—that is the ratio of orders to overall site visits—is 2.9 percent. While your mileage may vary, you still don’t know anything about what journey the visitor needs to take to convert to a buyer. There is so much that can happen between arriving on a site and checkout: visitors may browse, sign up for a newsletter or send a product to a friend. Each action reveals something about conversion. Merchants that study conversion rates in isolation, without understanding the path to conversion, have binary data—buy or no buy—but no idea as to why or when. As Forrester explains it, “[Conversion rate] is seriously flawed in that it does not lead directly to any kind of tangible improvement in a Web site’s customer experience. It’s too ambiguous and too binary—it’s either ‘up’ or ‘down’—to be directly relevant.”
Similarly, the abandonment rate is informative but not necessarily insightful. The abandonment rate measures the number of shoppers who put items in a shopping cart but fail to complete the transaction. The abandonment rate measures a certain aspect of site performance, but focusing on abandonment rate alone is misleading. You also need to look at visitor-to-cart metrics to determine the browseability of your site. Abandonment only measures activity once the visitor actually puts something in her cart.
If the metric for the conversion rate is too broad to produce actionable results, the abandonment rate is too narrow to describe the customer experience. Visitors abandon carts for all kinds of reasons. The path to checkout is fraught with pitfalls that are unique to the purchasing process. For example, the biggest reason for cart abandonment, according to Forrester, is higher-than-expected shipping costs. Also, Consumer Reports/WebWatch reports that 25 percent of online shoppers are concerned about personal security and sharing their financial information online. The numbers become more meaningful if you know why visitors abandon shopping carts. If you know why, then you can remove the obstacles to conversion.
Applying Metrics That Matter
Rather than focusing on metrics that are too broad or too narrow to be meaningful, focus on measurements that indicate behavior at key junctures on the path to purchase. Identify metrics that measure the level of engagement—the indicators that show when visitors become browsers and browsers become buyers. To illustrate, let me share some actionable indicators we took from our Web analytics data to compile the MarketLive Performance Index.
• “One-and-out” visitors. One-and-out indicates the number of visitors who hit one page and then leave the site without exploring or buying. One-and-out is an indicator of whether you’re meeting visitor expectations with regard to site and brand visibility and credibility. This is what I call the four-second rule—if they stay longer than four seconds, you’ve got them. You want to reduce your one-and-out rate.
Our findings show that 25 percent of one-and-out visitors hit the homepage, which is a good measure of whether your homepage has a healthy mix of offers and merchandise to appeal to a diverse range of customer interests. We found that sites with an average of eight different merchandising elements on the homepage had the lowest one-and-out rates and drove more than three times the conversions. We also found a correlation between one-and-out rates and visitor acquisition. The highest number of single page visits come from search engine traffic, so the search terms clearly need to match the customer expectation, which means pointing to specific product pages and using prominent merchandising and directive text.
• Cart Open Rate. Another common metric is the “cart open rate” or ratio of number of visitors to number of carts opened. If you’re only comparing the number of carts opened to completed orders, you’re missing the point. The cart open rate is the purest indicator of the power of your site’s content and merchandising.
Drilling down to find out what prompts visitors to open a cart is key. Site search, for example, prompts a high percentage of open carts (23 percent for online stores and 47 percent for catalogers, according to our Index), but many shoppers drop out before buying.
• Search-to-Cart. Studying search-to-cart statistics for specific search terms will show you where to focus merchandising.
• Product Page Visits to Carts. The ratio of product page visits to carts also shows if your pages have the “oomph” to drive conversions, or whether you need to revise the graphics or text, or give more prominence to your “add-to-cart” button.
• Checkout Abandonment. Cart open rate is an indicator of customer engagement, but to understand why shoppers complete a transaction, focus on the checkout abandonment rate—those customers who start the checkout process but never complete the order. Our study shows that, on average, about half of visitors abandon during checkout. If you start to see a high checkout abandonment rate, look for problems with your checkout process, such as forced registration, surprise shipping costs or just too many steps to complete the purchase.
As you can see, using analytics effectively doesn’t have to be overwhelming. All you have to do is distill your data and consider how to apply the findings in a practical way. Consider how analytic measurements reflect the level of visitor engagement.
So keep monitoring, testing and assessing, but always with an eye toward how you measure customer engagement. If you can use analytics to better understand the journey your customers make from when they enter the site to when they buy your products, you’ll be better able to remove any obstacles you see in their path.
Ken Burke is founder and CEO of MarketLive, an e-commerce technology services provider based in Petaluma, Calif. He can be reached by e-mail at firstname.lastname@example.org.