Coming soon to an online retailer near you: CCOMS

"It's you!" How many times have shoppers heard heard that expression? While shopping online, however, customers are pretty much alone in the store. The tools currently available to retailers are limited in their ability to speak to individual consumers. One day soon, however, your computer could become your personal shopper. A W. P. Carey School of Business marketing professor is working with engineering faculty to develop a system of software that will match products available for sale to the tastes of individual shopppers. The Customer Centric Order Management System (CCOMS) now under development would match your buying pattern with the stock available, allowing the web  site or telephone salesperson to offer you merchandise that you are likely to buy - just like the clerks in a brick and mortor store. 

Customer relationship management software allows retailers to collect information about what goes on "out front" — in the shopping arena where customers make their decisions.

CRM enables the assembly of large databases of information as detailed as how many seconds a shopper spends examining each item for sale on a Web site, which blouses she looks at, what color and style the blouses are, and on average what she spends per item, per total purchase and per year at the retailers' site.

Those same retailers also maintain a wealth of information about the workings backstage. Supply chain tools collects detailed information about inventory. Retailers know what blouses are at the warehouse and how many are hanging on a rack in a store, which stores are selling more of which blouses in which colors, which blouses aren't selling and which have the highest profit margin per item.

But retailers face challenges finding those types of specific information when they need it. It's much easier to gather and store information than it is to retrieve it, and harder still to use it quickly enough to drive sales.

If only retailers could take the two back-end systems and mesh them into a program that paves the way for personalized customer service while also maximizing profits.

Rajiv Sinha, a marketing professor at the W. P. Carey School of Business, started to think about this issue when he was doing his own online shopping.

"I noticed that two very different consumers visiting the same Web site would be presented with the same page," Sinha said. He wondered why the pages couldn't be individualized or tailored to the needs of different market segments.

That thought led to a research collaboration with Arizona State University engineering faculty as well as doctoral candidate Suraj Mohandas. Together they are developing a Customer Centric Order Management System. The system uses complex algorithms that allow three component parts to work together to customize the shopping experience benefiting both customers and retailers.

The first component is the customer knowledge base — the information that is collected about the customer's behavior and choices. This is the part that suggests alternative or complimentary products to shoppers. The second component runs mathematical formulas seamlessly in the background which are used to identify a "bundle" of products or a complete outfit tailored to the shoppers' taste. The shopper may be enticed with the information that the bundle costs less than the sum of its pieces purchased separately.

While others have tried to address this problem, Sinha said, they have done it through collaborative filtering programs. CCOMS uses an algorithm that can gather together the back-end customer and inventory information and present a completed, priced bundle.

According to Sinha, CCOMS advances the strategic use of customer and inventory data to the next generation. Retailers such as Amazon.com already use collaborative filtering to inform customers that "other customers like you" also bought gray slacks to go with that blouse, or that "other customers who bought this CD also bought CD No. 2." However, instead of relying on the behavior of "other customers like you," this enables retailers to make recommendations that satisfy individual customer tastes and increase the value of the transaction.

Sinha explained that CCOMS allows retailers to access information about an individual customer's taste, and then match it to stock available in the warehouse. If the blouse the customer wants is out of stock, CCOMS will identify other blouses that match her taste, offering first the blouse which is overstocked or has a higher profit margin. A coupon could be added to the mix as a gesture of goodwill because the first-choice blouse is not available.

All of this will happen while the customer shops, either online or on the phone.

"You're enhancing customer satisfaction," Sinha said. "The customer should be very happy because you're offering something she possibly wants but was unaware of because it was not present in her particular catalog. At the same time the retailer is moving inventory and doing it in a profitable way."

Making CCOMS work will be no mean feat, Sinha said. "The model is very technical. We're doing a lot of integer programming, it's a huge project," he said. The code is written and researchers are in the testing phase. They hope to have the testing done by the end of the summer.

Aberdeen retail CRM analyst Paula Rosenblum said the system was "very ambitious and a great thing for certain segments if they can make it work." However, she thinks the system will face challenges that have nothing to do with number crunching and everything to do with people.

The first is that it may be difficult to gather enough data on individual customer to know his preferences, especially for customers who are not regular shoppers at one retailer — problems Sinha said can be addressed by asking questions. However, Rosenblum points out that "everyone has a tipping point" at which they feel that their privacy is being invaded. Pointing out to customers just how much you know about them could make some of them uncomfortable. Still, she admits "there will be a certain customer who will appreciate that type of attention."

Sinha said he is not worried about privacy issues because most of the information the system takes into account is already being collected — it's just not being used to full advantage. Customers would have to register or create some kind of account, but the initial setup will be a guided, user-friendly process.

Although gathering customer data is a challenging task, especially for new customers, this problem can often be addressed by assigning the customer or client to previously established market segments — thus eliminating a huge portion of the questions that would normally be required for establishing their preferences, Sinha said. This scenario is more easily implemented in the B2B environment where the business requirements of companies in the same segment are similar. By sharing information, companies can sort their customers by preferences. The full profile needed top make CCOMS work would then be just a few questions away.

"It is our belief that it is the customer's shopping experience that encourages further business and not the actual sale," Sinha wrote. "The questions remain: What makes a customer feel like he got exactly what he was looking for? How can a company know their customer better to better meet their individual expectations?"

This, combined with maximizing profits and inventory control, is exactly what the Customer Centric Order Management System aims to do.