Instant Feedback: New database software offers online cross-sell opportunities

FOR THOSE READERS who think quick turnaround on the Web ends with an automated e-mail response, take note: In this year’s software review, DIRECT looks at two engines that collect data from online visitors and then offer product recommendations.

Precision Choice/Precision Insight

Online Insight Inc. thinks of its products as “virtual sales agents.” These sales agents lead potential customers through a series of exploratory questions regarding their needs, and then advise them accordingly.

The Atlanta-based company’s flagship product is Precision Choice. When customers enter a Web site using Precision Choice, they are asked a series of questions concerning their needs and preferences.

Using this information, marketers gain insight into how customers arrive at a specific decision, as opposed to relying only on past purchase events. Clickstream data does not reveal why customers looked at what they did or why they chose one product over another, notes Ina Beindorff, Online Insight’s chief marketing officer.

The software then scours the marketer’s product database and, through Precision Insight – the company’s recommendation engine – creates several hypothetical offers with different price points.

For example, a marketer offering financial services such as mutual funds can offer no-load funds vs. those with five-year returns, or showcase those that have been ranked highly by independent organizations.

Customers are then presented with either a one-stop suite of products tailored to their expressed preferences, or a top-10 list of products ranked by the weights the site visitors have assigned to various attributes.

One user of Online Insight’s software – financial products retailer YouDecide.com (www.you decide.com) – offers site visitors the option of going to the standard menu of products or taking advantage of the virtual sales agent. Forty percent say they want help, says Gary Snowberger, vice president of product development and alliances.

Of those requesting assistance, 80% complete the eight- to 10- screen interview process, which focuses on their preferences regarding down payments, risk characteristics and options as relevant to the product line they are investigating.

Although YouDecide.com does not provide this option, the newest version of Precision Insight allows marketers to build recommendations when the consumer stops offering information, even if the interview process has not yet been completed.

Customers believe that online sales agents are more likely than live agents to offer the best rates available, says Snowberger. As a result, commodity products such as term life or property/casualty insurance are among those most often quoted online.

But the biggest upside to the virtual sales agent is the ability to sell across YouDecide.com’s 18 product lines.

Live agents focus on closing a sale: The recommendation engine is programmed (if the consumer gives appropriate information) to offer additional financial products such as home equity or automobile finance loans, Snowberger adds.

YouDecide.com’s benchmark is a minimum of 1.7 product quotes in front of every site visitor, a figure that is not usually seen in the offline sales world.

Predictive Response

The flip side of making “live” recommendations based on expressed preferences is looking at shopping carts and making recommendations based on selected products.

Predictive Response, the first module in Delray Beach, FL-based Smith-Gardner & Associates’ Enterprise Marketing Automation Suite, uses collaborative filtering to make ancillary product recommendations.

Reporting functions allow marketers to quantify incremental revenue gains and cross-sell success levels. In a nod to integrated marketing, one feature provides suggestions regarding layout and configuration of print-based promotions based on complementary product sales.

Predictive Response can be tied in with inventory control software, which will prevent marketers from promoting items that are out of stock or have been discontinued. Alternatively, this feature can be used to promote clearance items.

The module’s capabilities include gift recommendation capabilities, suggestions for repeat customers visiting a marketer’s Web site and a cross-selling feature that makes suggestions as products are placed in a virtual shopping cart. This last feature fuels much of Predictive Response’s use at ClubMac, an online computer products retailer (www. clubmac.com).

“My sales people are trained to recognize what [customers] need,” says ClubMac president Mike McNeill. “The predictive engine is making the same type of recommendation.”

As customers add items to their shopping carts, the recommendations increase as well. These can be based on whatother customers buying similar packages have purchased or pre-set conditions, such as price point. For example, someone buying a $20,000 computer should receive a suggested add-on that is significantly more expensive than a few reams of paper.

The software can even be set to automatically recommend ancillary items customers don’t usually think of buying when purchasing online. Since implementing Predictive Response, ClubMac has seen a 20% increase in total line items per order – and has created a new market for at least one product category.

Unlike printers or serial cables, customers do not think to purchase warranties as part of their computer packages, says McNeill. Warranties are something call center representatives are trained to offer, and approximately 35% of all computer systems sold via inbound telesales include them. Predictive Response offers those to any customer purchasing a new computer system. Currently 15% of all online computer sales now include them, compared with virtually none before.

McNeill hopes the next version of Predictive Response will incorporate top-10 lists of recommended purchases, and a feature that will automatically track consumable purchases such as ink and paper and make reorder suggestions based on elapsed time between orders.