Who Knew?
What your web visitors are trying to tell you
By Mary Wagner
The Egyptians had the Rosetta Stone, the Allies had Enigma and today’s web site operators have web site analytics. E-retailers have access to a mountain of site data; uninterpreted, it could remain as meaningless as hieroglyphics or undeciphered code. But over the past few years, web site analytics software and services have stepped up to take on the role of cryptographer.
For merchants, analytics applications are moving beyond log files that simply record traffic to a new generation of more sophisticated implementations that use data tags to track the customer’s path deep within the site. And meaning has emerged from those measurements. Data become useable business intelligence, forming patterns that help drive merchandising decisions, prioritize the placement of site features, and guide marketing expenditures. “If you can point to a conversion rate of 4%, that’s nice to know, but it’s not actionable,” says Brent Hieggleke, director of WebTrends marketing at analytics provider NetIQ Corp. “You might want to get to 5% next quarter, but what action do you take to do that?”
Web analytics provide such answers, and hosted analytics solutions that lower the initial cost of use to under $100,000 are expanding their use. “We’ve seen an evolution in what sites use analytics for,” says Steve O’Brien, vice president of sales and marketing at analytics service provider Fireclick Inc. “It started with measuring conversion rates, then traffic. But most people today are using it for tracking not just the conversion rate, but the revenue associated with specific campaigns.”
And it doesn’t stop with tracking campaigns by revenue—today’s analytic tools go beyond campaign revenue to track campaign profitability. Viewed collectively, the experiences of the retail analytics user who discovers that site traffic doesn’t equal revenue—or that switching to the right keywords at the right price does—are more than individual stories. With a window on the activity of many e-retailers, the providers of analytics services see trends in the stumbling blocks and the light bulb moments that repeat themselves among their clients as they use analytics and their learning curve goes up.
Gleaned from the files of providers are some of the discoveries typically made by online retailers as they dig into analytic data.
When you’re paying too much for traffic
Retailers pay big for customer referrals from sources including business partners, search engines and keywords, and others; analytics can pin down just how productive these relationships are.
One WebSideStory retail client who spent significant time and resources in working with various business partners to drive customers to its online catalog got a wakeup call when it started tracking referrals with HitBox. Analytics showed that a partner who had charged the retailer a premium on the claim it was delivering a large amount of highly qualified traffic actually delivered very little traffic. Armed with the numbers, the retailer renegotiated its contract with the vendor on more favorable terms. The savings from the renegotiation paid for the retailer’s first year of HitBox service.
“That’s a 100% return on investment from a single good decision that was made possible through accurate measurement,” says Eric Peterson, senior e-business analyst.
Which online marketing dollars
are going down the tubes
“The first thing that surprises people when they apply analytics is how much of their marketing is being wasted,” says Chi-Hua Chien, director of marketing at analytics provider Coremetrics Inc. “Keywords are now all the rage. With keywords, you’re paying for clicks. Show not only the number of clicks that a keyword generates, but also the amount of revenue and then profit on a margin contribution basis, and it can be shocking for an e-retailer to see that the 10,000 clicks they’ve just paid $1.25 each for brought in revenue of $100. They’ve just paid about $12,000 for virtually no revenue and even less profit.”
Chien says every retailer who applies analytics to keywords has this experience on one scale or another. But while there’s an old saw in marketing circles that marketers know half of their budget is being wasted, but not which half, that’s precisely where analytics can shed light. One Coremetrics e-retailer client using CoreMetrics’ MarketForce product reduced customer acquisition cost by 79% with tighter tracking of which keywords produced conversions and which didn’t. A seller of big-ticket items, it had discovered that its cost of customer acquisition in specific keyword categories was extremely high—close to $100 per purchase. Using the provider’s analytics tool to track each keyword’s contributions, it reduced its bids on lower-performing keywords and upped its bid on top-performing keywords to a No. 1 or 2 spot on the top 10. That resulted in reducing customer acquisition cost in those categories by 79%, while driving through more revenue to nearly double the ROI on keyword investment, says Chien.
When
the problem isn’t what you think it is
E-retailers may think they know where their challenges lie, but analytics often shown otherwise and that they need to shift focus to maximize ROI. In the case of one Fireclick client, analytics revealed that simply investing in a new shopping tool would deliver no ROI unless the retailer also figured out how to make shoppers use it.
Fireclick’s client, a large online retailer, was contemplating adding rotational 3-D capacity for viewing selected products. Before it invested to add the necessary software and take the extra product photos, it collected and compared data from three customer segments. One group had access to the tool and used it, the second had access to the tool but did not use it, and the third didn’t have access to the tool. Fireclick’s Netflame analytics product found that the conversion rate between those who had access to the tool and didn’t use it and those without access to the tool was identical. The conversion rate among those who had access to the tool and did use it, however, was double the rate of the others.
“Just by making the feature available, nothing would happen to the conversion rate, but if they could somehow convince people to click on the 3-D button, they could double their conversion rate on that item,” says Steve O’Brien, vice president of sales and marketing. “So if they spend time and money, it shouldn’t be to make sure that the button is on every page, but on making sure that people use it where they do find it. It. They never would have known that without analytics.”
Why
registration at checkout can be poison
Analytics providers say too many online merchants place registration at checkout because they associate this step with closing a purchase. But that can annoy shoppers. Eric Peterson, senior e-business analyst at analytics provider WebSideStory Inc., offers an analogy: “If you filled your shopping cart at a grocery store, waited in line at checkout, and then the cashier made you fill out a form and wait to get approval from the manager before you could put your groceries on the belt, how often would you buy from that store?” he says.
Relocating registration to elsewhere in the process or even making it optional has paid off, as analytics have demonstrated for retail users of WebSideStory’s HitBox analytics tools. One e-retailer moved optional registration away from checkout and saw conversion rates rise. Its rate of home page visitors who became buyers more than doubled to 2.2%, while the conversion rate as measured from the starting point of the checkout process tripled from about 11% to 33%.
“We don’t fault online merchants for wanting to collect registration information,
but when customers go to checkout, they expect an order summary and questions
about billing and shipping. They don’t expect to have to log in then, or create
a record if they don’t have one,” Peterson says. “Registration can be very elegantly
moved out of that place and away from being an impediment."
How
campaigns may be doing better than you think
At a time when e-retailers are tracking the return on every marketing dollar, the application of analytics often shows where marketing dollars are being wasted. But it also can show the opposite, as Brooklyn, N.Y.-based retailer Etronics found. When it applied a recently developed module from Fireclick’s Netflame service to its online campaigns, it found it was getting twice the conversions on the campaigns it had initially believed.
The new module captured deferred conversions, in addition to the direct conversions already tracked by the service. “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.
Merchants who send out mass e-mails or buy keywords typically track revenue by purchases that occur when a shopper clicks to the site from the promotion and immediately makes a purchase. But that doesn’t capture shoppers who clicked through, didn’t buy then, but bought later, a common occurrence with higher-ticket or more considered purchases.
“The deferred conversion feature sees a cookie that says a visitor had originally been brought to the site by this keyword or that e-mail, and it credits that campaign with the conversion,” says Fireclick’s O’Brien. The feature is set by default to recognize site visitors originally brought in from a promotion for seven days after they visit, but the time period can be set at whatever a merchant wants, he adds.
For Etronics, which sells consumer electronics items, a higher-ticket product category, the tool proved essential in judging ROI from its marketing campaigns, Balser says. It swung the company’s decision to stay with some online ad campaigns it had considered borderline by showing that, including deferred conversions, the campaigns were actually performing better than initially believed. “People do come back to the site to buy in a second, third or fourth session,” he adds. Fireclick estimates that for a given marketing campaign, as many as half of conversions may occur during a deferred session.
If customer demand is hiding
Increasingly, e-retailers understand that the Internet is not just a transactional channel but a market research channel as well. What shoppers look for but don’t find on retail sites may be just as telling as the sales numbers on products they do buy, if those unsuccessful searches are tracked and analyzed.
A major multi-channel e-retailer that didn’t carry furniture online or in its stores discovered through Coremetrics’ analytic tools that a large number of online shoppers were nevertheless searching for it on the site. Their site searches under “furniture” and associated keywords either came up empty or delivered results with low relevance, such as “furniture polish.”
“This was a shocker, because you don’t generally want to go to a web site, buy something that weighs 200 to 300 pounds, and have it shipped. But this retailer found that a lot of people wanted to buy furniture from them,” says Coremetrics’ Chi-Hua Chien. Based on the search patterns revealed by analytics, the retailer used its web site to launch a new furniture product line, experienced high conversion rates online, and went on to test response to the new line on a regional basis in stores. “Companies are finding that they shouldn’t view their web site only as a store. It lets then do much more,” Chien says.
When less is more
It’s not uncommon for an e-retailer to have as many as 40 or 50 links on a home page, including those on drop-down menus. Net IQ Corp.’s Brent Hieggleke says he’s seen clients with 100 or more links off the home page. Retailers’ thinking is that the home page is the first and possibly the only opportunity to win a visitor’s business, but analytics have shown that putting too much on the home page has the opposite effect.
Analytics showed Hieggleke’s client GeoPassage, an online travel company, that its home page was one of the top exit points for visitors. That finding indicated that the page wasn’t performing its function of drawing people deeper into the site, and in fact, was overwhelming them with too many choices that caused them to flee.
“The company needed to understand that while someone coming to a retail site is going there to purchase something, they are probably not going to purchase anything on their first visit. The scenario for a first-time visitor is very different than for a repeat visitor. You have to make sure that you design the home page for both scenarios: the first-time visitor and the one who is returning to buy something quickly,” Hieggleke says.
The key is to arrange the page so both customer segments can immediately find
what’s relevant to them, he adds. For example, that could mean a concierge button
targeting new visitors that unfolds content in a logical progression instead
of displaying everything the visitor needs to know all at once. For returning
customers, it could mean navigation that highlights a quick path to checkout
so they can skip what they don’t need to see. By simplifying and reorganizing
its home page, and testing successive iterations, GeoPassage arrived at a page
design that featured only the top destination countries. In short order, this
increased the number of return visitors by 30% and eliminated the home page
from the list of the site’s top exit points, increasing the average length of
time visitors spend on the page by 33%.
Where
the bottleneck is
There’s at least one choking point in every retailer’s online shopping process—often more, according to Brent Hieggleke, director of WebTrends marketing at NetIQ Corp. “There’s always a spot in the purchase process that drops out more customers than other spots. Once you identify it, you know where to attack,” he says.
One of Hieggleke’s e-retailer clients discovered that its shopping process
was taking too many clicks and too much time. It identified exactly where in
the process it was losing shoppers, then reconfigured the steps to reduce clicks
and shave off about 30 seconds. Conversions jumped 25%—from 4% of shoppers to
5%. “That’s like free money,” Hieggleke says. “They didn’t have to spend anything
or drive any new marketing campaigns to get it. They just had to get more efficient.”
mary@verticalwebmedia.com
Click
here for the Guide to Web Analytics Solutions 
|
