Automation can deliver improved marketing results once e-retailers build bridges between data silos.
PureFormulas.com carries about 14,000 SKUs, which means the vitamin and health supplements e-retailer has a hard time making sure its marketing reflects the most recent prices, promotions and in-stock items. That's why the web merchant recently introduced automation into its paid search campaign.
Using a ChannelAdvisor Corp. tool called Inventory Driven Search, PureFormulas ensures that its paid search ads reflect the current product data in the feed it provides to ChannelAdvisor, an e-commerce marketing technology and services firm. The tool automatically updates ads to reflect the latest prices and promotions. And if an item is out of stock, such as the 255-gram size of the Vital Scoop energy shake, the technology will substitute an ad for the 510-gram size.
In addition, the ChannelAdvisor tool suggests keywords specific to each item, such as "energy shake vanilla 255 grams." Because those phrases are so specific, there is little competition for them and they cost relatively little. At the same time, a consumer who searches for that phrase likely is intent on buying the item, thus those phrases often convert at higher rates than other keywords.
PureFormulas, which began using the tool about five months ago, credits it with helping to decrease the retailer's ad spend by 20% and increase search revenue by 30%, says Jose Prendes, the company's founder and president. "If you're running a paid search campaign online, this is a no-brainer," he says.
PureFormulas' paid search effort illustrates how web retailers can squeeze more profit from marketing efforts if they can use the data they have about products and customers to automatically trigger e-mail messages and ads likely to be relevant to customers and that take into account a retailer's current inventory.
The challenge is to get disparate systems to work together, such as the analytics technology that tracks an e-retail site's most popular products with an e-mail marketing system. One indication that such automation is in its infancy comes from digital marketing services vendor Webtrends Inc., which estimates no more than 10% of retailers integrate analytics with marketing systems.
The promise is that e-retailers can market more effectively without human fingers constantly pushing buttons. Last month's release of the ChannelAdvisor Inventory Driven Search tool is one example of vendors making available more tools that can automate marketing, particularly e-mail, online display ads and paid search.
For these tools to work, a retailer must have good data about its inventory, and about its customers and site visitors who may become customers. The most important customer information retailers need to collect concerns how frequently a shopper visits and buys, the last time she bought and the value of that purchase, says Ellen Duffie-Fritz, head of marketing strategy and analytics for TrueAction, the marketing services division of e-commerce technology and services provider GSI Commerce Inc. Such data enable a retailer to send in a timely fashion automatic e-mails, or serve up relevant display ads, to consumers in the mood to shop and buy; the messages could include discounts for loyal customers who might be prodded to buy a bit more, or reminders that items were left in shopping carts.
Web retailers usually have an abundance of such data. "They are sitting on a pot of gold," Duffie-Fritz says. However, she adds, retailers don't always have the data in a useful form, which can mean, for instance, that a retailer may have to add additional data to code product feeds, or that data must be shared between departments in new ways.
A big challenge Charming Shoppes Inc. faced last year when preparing to test retargeted display ads was getting its data into shape so that vendors competing for the retailer's business could use it, says Ken Mowry, senior vice president of marketing.
The women's apparel retailer, which operates Lane Bryant and other brands, had to update the coding on its product feeds and fix other areas, such as making sure that a promotional price actually paid for an item was not mistakenly inflated to its full price on back-end reporting systems; such a mistake could affect the retailer's ability to measure the returns from marketing programs. "Accuracy of the data is critical," Mowry says. And there was a cost associated with that data preparation. "Anytime you touch code on the page, you are involving the technology team," he adds.
Tagging, planning and budgeting for the Charming Shoppes automation effort took about two months, Mowry says.
Similarly, CafePress.com, which sells customized merchandise, underwent four months of prep work to get its product data into shape so that a vendor could send automated offers to good customers, says Sumant Sridharan, vice president of online acquisition. For CafePress, that meant figuring out how to put customers in some 20 segments based on location, gender, purchase amount, type of purchase and other factors, which then leads to post-purchase e-mails that offer discounts and offers based on those variables.
First, the company had to decide what types of segments to divide its customers into, and then arrange the data in such a way to populate those segments. For instance, CafePress wanted to identify customers who made relatively large purchases over certain periods of time, which meant having its development team go into the guts of its customer history data to sort and combine the necessary information so that CafePress could determine who warrants juicier discount offers.
All that work can be worth it. Based on client experiences, Duffie-Fritz says, automatically generated display ads can boost conversions by up to 30%, while automated e-mail messages triggered by consumer behavior can be as much as five times more effective than other e-mails.