Web Site Analytics
Show me the money
ROI is the goal of almost every move marketers make on the web—every merchandising decision, keyword purchase, technology addition and page design change.
Not long ago, online merchants simply guessed about the return being produced by any of those elements and considered themselves on target if log files showed overall traffic and sales were hitting projections.
But today, analytics is a hot topic in online retail as analytics vendors have figured out how to tag data and slice it into ever-finer segments that reveal patterns in visitors’ paths to a site and movements through it. And not only have analytics vendors determined how to dig out data in ways that are meaningful to site operators, but they’ve also started pricing such services to put them within the reach of more retailers. More expensive and cumbersome enterprise-scaled analytics software installations have largely given way to less-expensive hosted applications that online merchants buy from vendors as a service.
No more software discussions
“The ASP, client-side tag model has won,” says Jupiter Research Inc. analyst Matthew Berk, referring to both the hosted application model and technology that tags code to web pages so as to send information from the user’s browser to a central source whenever the page is viewed. “Only in rare circumstances might you want to run software. A year ago, all of the calls we got about analytics from companies were from those who were evaluating different software. Only rarely now do we get questions about the difference between outsourcing and running software.”
One of the biggest differences is cost. Hosted solutions now undercut the price of buying software licenses by tens of thousands of dollars and more. “CFOs saw they could spend a quarter of a million to get a web analytics solution or they could look to an ASP solution with a service level agreement they could cancel if they weren’t getting the results they wanted,” says analyst Guy Creese of the Aberdeen Group.
By zeroing in on smaller, more sharply defined segments of customer behavior, analytics have also given site managers charged with improvement something to shoot for. Data from analytics are breaking global objectives such as boosting revenue and profit into smaller, more manageable—and measurable—tasks. Some of the most successful retailers approach their use of analytics in a classic experimental model, forming a hypothesis, grounding it in data, making a site modification and then measuring the change.
Doug Greene, chief technology officer and chief marketing officer for Overstock.com, says his company’s ability to quantify the effects of changes to the site is biggest benefit of analytics. “Given all the noise, it is impossible to quantify the results of small changes to pages, and whether they help or hurt conversion, unless you have this kind of tool,” he says.
New landing pages
Flower site Proflowers.com recently put analytic data to work in a page redesign, testing visitor response to different page elements and grouping the winners in a new test landing page. Over the next six weeks, the new page containing all the top-performing elements drove a 10% to 15% increase in conversions, says web strategist Richie Hannah.
“We’ve had opinions of what we thought would help or hurt conversions, but we’ve seen time and time again that what seems to make logical sense isn’t necessarily true,” Hannah says.
As Proflowers’ experience shows, the application of analytics is impacting page design. By pinpointing bottlenecks in the checkout process that cause dispersion among shoppers who’ve already initiated buying, for example, some online merchants that discovered their checkout was too long or had too many steps have boosted conversions by making it more streamlined. One online seller of travel packages used analytics to find that it was overwhelming site visitors with too many choices on its front page and actually causing them to leave the site. By simplifying the home page to feature offers on only the top-most visited destinations, the site increased by almost 33% the average time visitors spent there, according to Brent Hieggelke, director of WebTrends marketing at Net IQ Corp., the site’s analytics provider.
One client of analytics provider Fireclick Inc. used the tools to discover that his online marketing campaigns were actually doing better than expected. By tracking deferred conversions—when a site visitor brought in by an online campaign bought not then, but on a subsequent visit—the web site of consumer electronics retailer Etronics found that its online campaigns were performing better than it had initially believed. “Etronics.com is making twice as much money as we thought we were from our marketing and we had no way of knowing this before,” says Etronics vice president and owner Mayer Balser.
Richer reporting
Recent generations of analytics tools have sought to automate data collection as much as possible, on their developers’ assumption that this would provide the greatest practical utility to the marketers who needed to use the reports. But some marketers want data sets that don’t fall within pre-set reporting formats, as analytics providers such as Omniture Inc. learned. “Previously, we thought the only way people would be willing to track campaigns was if SiteCatalyst (Omniture’s analytics product) could automatically detect campaigns and automatically provide reporting on how those campaigns were driving results. We still do that, but for those that are willing to make that additional data entry into SiteCatalyst, we can enhance the richness of the reporting they get considerably,” says Brett Error, Omniture’s CTO.
Toward that end, in addition to the standard reporting format, Omniture’s 9.5 version of SiteCatalyst, the latest release, offers marketers the option of classifying, grouping and comparing campaign creative elements within campaigns according to criteria of their own choosing. The SiteCatalyst campaign management center centralizes reporting on all campaign data. Marketers can classify the creative elements of a campaign in various ways, such as, for example the type of media: flash ad, static or animated gif, link, or direct e-mail. “The end result is that they can get reporting on all of the classifications. They can see if their jpeg ads are performing better than their flash ads, for example. And they can add in all the standard metrics, so they might find, for example, that while flash ads were better at driving people to the site, they are for whatever reason less effective at driving conversions,” Error says.
Another online marketing realm on which analytics have shed particular light and even caused marketers to shift spending is in the area of keywords. With search pegged by some industry analysts as the third largest online market, it’s getting a bigger share of online marketer’s budgets. Increasingly sophisticated analytics can help validate that it’s money well spent — or not.
One way is tracking keyword results beyond simple click-throughs on to purchases, profitability, and ROI. “Some competitive keywords are really expensive when you look at their cost per click and cost per visit,” says Hieggelke. “But taking it thorugh click-through is just the beginning. You might be getting a lot of visits from a certain keyword, but they might just be tire-kickers, or shoppers who check your price pages and leave immediately, which tells you that they are price-sensitive, and that might not be your target audience. You have to judge the quality of that traffic.”
Net IQ’s WebTrends now not only tracks keywords and other campaigns such as banner ads and e-mail beyond click-throughs to conversion, revenue and lifetime value, but it recently added functionality that distinguishes between pay-for-performance and natural, or algorithm-based, search results.
That’s important because search engine users are becoming savvier about distinguishing between the two. Both types are delivering results for marketers and most search experts recommend a combination strategy that uses both. But to apportion both resources and time most effectively between the two, marketers need to measure the performance they’re getting out of each.
The coming rationality
“You have certain words on which you want to be ranked high in natural search results—but if you aren’t, you can get there by using paid ads. We’re still in a bit of a gold rush, with people buying thousands of keywords because they just don’t know. But people will become very selective about what keywords they pay for,” says Hieggelke. “Once they start measuring rationally with tools like the ones we provide, they’ll realize that there are certain words they need to pay for and others for which they should just cut bait and move on.”
Analytics have advanced retailers’ ability to use the power of the web to gather customer data. Retailers’ growing challenge today is not so much getting customer information as using it to form action plans that support business goals. That’s a sticking point for many—one reason that, despite a wealth of data potentially available from analytics tools, a whopping 60% of merchants who do measure search engine marketing campaign response measure only click-through rates and general traffic, according to a survey by NetIQ.
That sets the stage for analytics firms serving retail clients to help them make the leap by showing them how to use those results. And that, in effect, expands their role from technology provider to service provider. The top analytics providers will start to work in partnership with retailers to share knowledge and identify where and to start leveraging the data, says Berk, who adds, “We’ll see a lot more of that educational component in the coming year.”
