Best Websites

Since 1998 BestWebsites.com.my features thousands of best websites
in many categories of interest with descriptions/reviews given by leading
publications and webmasters.

What your web site can tell you

10 metrics to watch to ensure quality site performance
By Peter Lucas

When it comes to web analytics, there is no shortage of measurements retailers can take to understand how their site is performing. On average, most retailers track hundreds of metrics daily, weekly and monthly. All this data helps them understand how and why browsers convert to buyers, which portions of their site or promotions are working effectively and which are not.

Sifting through such vast amounts of information can be mind numbing. Even more daunting is knowing how much weight to put behind each measurement and how to put the data into context. In other words, it’s easy for retailers to identify which performance characteristics to track, but not necessarily how to interpret the resulting data and know what to do with it.

Retailers agree that knowing which metrics to track and how much weight to give each one comes down to a very simple premise: understanding what they want to accomplish when it comes to site performance.

“It’s not just a matter of knowing what visitors do on your site and how they behaved once they got there, but understanding how that information can be used to pinpoint problems, improve performance, even for what works, and generate more revenue,” says Richard Calentine, Internet analyst for Baseball Express Inc., which sells baseball equipment on the web.

While performance goals will vary by retailer, making a list and prioritizing goals can aid retailers tremendously when it comes to working with analytics vendors to identify the necessary metrics. Retailers can also be well served by regularly reviewing the order in which they prioritize their performance goals as they are likely to vary by season. Retailers must also keep in mind that too many metrics can lead to information overload, which slows response time to potential problems or in capitalizing further on a current success, resulting in missed sales.

Following are 10 key areas of performance measurement that most retailers have—or should have—at the top of their lists, and the insights the data gleaned from those metrics can provide.

1. Study Site Search Results—and Non-Results

Examining what shoppers are looking for—and finding and not finding—may seem like a no-brainer, but the truth is that the words consumers use to search for a product ebb and flow. As a result, retailers lose a lot of sales because consumers can’t find what they want based on use of a search word not in the retailer’s dictionary. One way to remedy the problem is to track searches that turn up no results, a.k.a. 0% search.

As part of a recent site relaunch, BaseballExpress.com overhauled its search dictionary to include weight differentials for bats, a description many customers used in their site searches. The payoff was about a 50% drop in 0% searches. “If you can’t get people to the product for which they are looking, they are not going to buy,” says Calentine, who adds he expects conversion rates on those searches to be up substantially as well.

Retailers can also use 0% searches to determine if they need to add a frequently searched product to their inventory or more simply, whether they need to add a new term to their search dictionary. Site search can even be tweaked to return like items, in lieu of not carrying the actual item, which increases the chances of keeping the customer on the site and converting the sale. The savviest retailers can cut deals with competitors to return a referral link for a non-inventoried item in exchange for a finders fee.

“Site search is also a merchandising tool,” says Warren Raisch, senior vice president of analytics vendor WebSideStory Inc. “People who use site search convert, on average, at a 3% higher rate than customers using standard navigation tools. There is no reason not to make use of the merchandising capabilities of site search.”

2. Optimize the Home Sweet Home Page

Possibly the most valuable piece of promotional real estate retailers own, the home page is not a place to let poor converting elements linger for long. Still, it’s not enough to identify a poor converting space—retailers are best served if they can determine the actual revenue of each element.

“That means looking beyond the click rate to see the average revenue on a per-click basis,” explains Matt Belkin, vice president, Best Practices Group for Omniture Inc. “Once retailers see that, they can surgically cut poor performing elements, figure out what lags and if it’s worth fixing, or figure out what works and promote it.”

These metrics can be especially helpful to retailers that stock seasonal items. HaleGroves.com wastes little time on certain varieties of fruit as they come to the end of their season and switches to Hot Sellers for the incoming season. The same is true for post-Christmas shopping when focus on the home page shifts from gift packages to seasonal varieties. “Home page optimization comes down to conversion rates,” says Paul Lazorisak, director of marketing for Hale Groves Inc., a multi-channel retailer of fruit. “You want elements and promotions that generate good click-through rates.”

3. Know What Works on Landing Pages

It makes no sense for retailers to pay attention to optimizing their home pages if they won’t create custom landing pages. More than half of customers will land at a retailer’s site either through a search engine, banner ad, e-mail promotion or affiliate marketing partner. It makes sense then for retailers to create landing pages designed to appeal to these customer’s preferences for product presentation and branding based on the vehicle used to transport them to the site.

Analytics can tell the retailer who comes directly into a landing page and who they are likely to be. “Landing pages ought to speak to the persona of the customer,” says WebSideStory’s Raisch. “When that is achieved, retailers can feed customers the right information at the right time. It’s about moving away from a feel of what the customer wants and using more science.”

BaseballExpress.com customizes the product on its landing pages linked to affiliates based on the audience attracted by the affiliate. “Several of our affiliate partners are Little League organizations, so we will customize the product selection for that segment,” says Baseball Express’s Calentine. “One of the reasons we redesigned our site was to be able to provide a more personalized experience.”

4. Use the Shopping Cart for More Than Just Checkout

The shopping cart is no longer just a vehicle to get customers to and through checkout; it is a shopping tool. If a consumer puts an item in a shopping cart, then takes it out, or worse, leaves it in the cart, but abandons the cart, odds are the retailer is not using the cart as shopping tool. A leading reason for cart abandonment is that consumers get distracted by jumping to other pages after clicking on a item in the cart for review.

Analytics can tell a retailer when that is happening. “Retailers need to examine the clickstreams within the cart to identify where people drop off, if they put a removed item in a wish list, and whether they come back if they jump to another page,” says Jane Paolucci, vice president of marketing for Coremetrics.

There is no reason to transport the customer from the cart when pop-ups can deliver the information on the product to be reviewed. The same concept can be applied when it comes to initiating checkout through the cart, only instead, shoppers are fed cross-sell or up-sell items that can be added to the cart without moving to another page.

Another key data point that can transform the cart into a shopping tool is keeping a running tally that includes tax and shipping costs to prevent sticker shock at checkout. Armed with this information, consumers can better manage the items in their cart if they are on a budget and complete a sale, rather than abort it later. “Sometimes consumers need a larger picture of the shopping experience to enhance it and increase conversion,” says Josh Manion, CEO of Stratigient. “Providing that can create a greater value for the customer with each visit.”

5. Look at the Look-to-Book Ratio

Because so many elements feed into the broad-based look-to-book metric, retailers are best served by examining trends and putting them into historical perspective. HaleGroves.com typically does a lot of business around the holidays. With Easter falling in April in 2006, as opposed to March in 2005, the company knows to temper its view of the browser-to-buyer ratio when comparing the weekly data from those two months year-to-year. The company also has a lot of browsers at the beginning of the Christmas shopping season and knows that conversion rates will pick up substantially beginning in late November.

“It’s important to step back and view visits in context,” Lazorisak says. “Some holidays fall on different weeks of the month each year and other periods are prone to high browser ratios. Data patterns provide insights and can make the difference between trusting your data and panicking.”

Data patterns are but one clue to improving look-to-book ratios. The number of times a consumer views a product before buying it, or not, can tell retailers such things as whether the item is over-priced compared to the competition, whether the Add to Cart button is visible or whether the customer got distracted after linking to a related page. Items with high view rates, but low add-to-cart rates are in need of close scrutiny from all angles.

“The aim is to understand the conversion process,” says Jason Palmer, vice president of products for WebTrends Inc. “Identifying customers that do a lot of research before buying can help retailers create a more engaging experience that converts a higher percentage of them faster.”

6. Understand How Shoppers View Cookies

Cookies are what make it possible to recognize a return customer, but with the proliferation of spyware applications on shoppers’ computers, having an analytics vendor attach a cookie on behalf of a retailer no longer makes sense. Spyware will automatically block third-party cookies, meaning that repeat customers lose all recognition of that status each time they return to your site.

Few repeat customers like losing bookmarked pages, wish lists or checkout data they thought was stored due to a blocked cookie. The solution is for the retailer to attach a first-party cookie. “As the use of spyware on computers proliferates, retailers are going to have a harder time retaining a memory of their customer’s visits,” says WebTrends’s Palmer.

DesignerLinensOutlet.com, which recently switched to a first-party cookie, reduced the number of rejected cookie attachments by 95%, while enjoying a 45% gain in return visitors. “It’s a huge benefit to be able to leverage our own cookie,” says Beverly Dantz, marketing specialist for Designer Linens Outlet Direct.

Dantz adds the new cookie has provided insights into how the company can spend marketing dollars more effectively.

7. Know Affiliate and Search Engine Marketing ROI

Retailers pay plenty for these services, so they’d better know the actual cost. That means accurately tracking how many times a customer comes to a site through the same affiliate or search engine and what she does once she gets there.

Too often, a customer shopping at an online store gets directed there through an affiliate only to be directed back to the same site through another affiliate when searching for a different item. “A lot of retailers are unaware they pay three to four times over to acquire a single customer through these channels,” says Omniture’s Belkin. “Retailers need to understand what customers buy once they have visited their site and move the relationship away from the affiliate to themselves.”

That means taking a close look at the affiliate through which the customer came to the site and the keywords entered into the search engine that landed them with the affiliate in the first place. Next, retailers can focus on pushing other items likely to be of interest to the customer via e-mail, if only to let the customer know they have an extensive and varied inventory.

HaleGroves.com, which tracks revenues per affiliate relationship as a way to maximize marketing dollars and control acquisition costs, has discovered that it is best to limit its affiliate marketing. “We know that when people first come to us, they usually are not ready to buy,” says Lazorisak. “Having a key measuring platform in place really helps us to know where best to spend our marketing dollars.”

8. Use Analytics to Make E-mail Marketing More Effective

Populating e-mail with content of importance to each customer is another no-brainer, but something that often gets short shrift because retailers tend to think they need to give away marketing dollars to make e-mail effective.

The whole aim is to drive traffic without increasing marketing costs through discounts. In some cases, information about a product previously viewed or notification that an item on a wish list is close to selling out or being discontinued, can spark a sale. Offers can be sweetened by playing up applicable promotions already going on, such as free shipping. Finding which promotions work best is achieved through A/B testing. “A/B testing lets you measure the effectiveness and revenue per acquisition to identify the best performing campaign before it rolls out,” says Lazorisak. “That’s very valuable.”

Analytics can also help determine the life span of a promotion by identifying the first hints of when it starts to lose momentum. At that point, retailers can opt to tweak it or prepare to immediately implement a new promotion once the existing one no longer generates worthwhile revenue.

9. Manage Customer Category Movement

Consumers may purchase from multiple categories, but there is no reason to let them run around a web site putting together an outfit when they don’t have to. Analytics help merchants identify buying patterns as customers move across categories, enabling retailers to more effectively cross-sell items.

Suggesting a ski outfit to a consumer who has put a ski jacket, gloves, and related accessories into a cart can not only increase the size of the sale, but result in a more satisfied customer. “Customers will buy in one category and often browse another during the same session, so there is a lot of opportunity for cross-selling,” says Belkin.

Retailers can also use analytics data to determine if a repeat visit is likely soon. By analyzing customer movement and behavior within its site, DesignerLinensOutlet.com determined that many customers were likely to make a return visit within a specified period. The retailer now sends special offers to entice customers with those characteristics back to the site sooner. “The patterns in customer behavior have helped overall marketing and revenues per marketing dollar,” says Designer Linens Direct’s Dantz.

It also helps to know your product, she adds. As a retailer specializing in closeout items, DesignerLinensOutlet.com rarely suggests items during a repeat visit due to its high turnover of inventory. “It’s pretty tough to recommend something after the first sale, because what made sense to cross-sell at the time may no longer be available at a later date,” adds Dantz.

10. Find Out Why Shoppers Are Not Checking Out

Nothing is more frustrating to retailers than getting a customer into checkout, only to have her abandon the process. Analytics can pinpoint where the customer dropped out. Once all the pieces are known, the retailer can determine whether the process is too long, too cumbersome or does not provide enough data at the point the customer appears ready to buy.

The last is critical, because information about return policies, date of shipping, and customer service contacts provide comfort for the customer. So does automatically populating data fields, such as mailing address, for repeat customers. “There are key elements customers need to feel comfortable with to get through checkout,” says Raisch. “Understanding these in detail can cut abandonment.”

Still, providing a comfort zone won’t drive checkout abandonment to zero, which is why some retailers watch checkout in real time. This technique is not so much for identifying potential problems, but more for saving sales. Capturing the data on what was in the customer’s cart at the time can open the door to a follow-up e-mail promotion.

For such follow-up programs to be successful, of course, the customer had to have provided an e-mail address before abandonment. Once in possession of that data, the retailer can send e-mails pitching discounts on items that were in the cart as a way to encourage a return visit. Even without an address, retailers can ID the customer by cookies on a return visit and offer a similar promotion then. “Knowing what’s in a cart at the time of abandonment is a great way to create targeted promotions,” says Coremetrics’ Paolucci.

Avoid the one-dimensional

Applying web site analytics is a complex undertaking. Understanding the results can be even more daunting. But that understanding is essential to a web site’s success. “The true value of analytics is understanding the insights provided by the numbers,” says DesignerLinensOutlet.com’s Dantz. “Assigning a value to each metric is one dimensional.”

And as retailers know, multi-dimensional thinking is the key to keeping their web site performing at the highest level. l

Peter Lucas is a Highland Park, Ill.-based freelance business writer.

How to hear what your web site is trying to say