Using data to build relationships

“To be successful now, in a difficult economic climate and going forward, companies must understand and respond to the shift from technology for information to technology for relationships.” Lars Nyberg, NCR Corporation chairman and CEO, made that statement in his opening remarks at the recent NCR Partners 2001 conference, data warehousing and CRM event in October in Orlando, Fla.

An overall message that “knowledge is the key to power, profits and safety” was directed to some 1,500 NCR Corporation customers and prospective customers attending about 170 conference sessions. NCR promised to show them how to better use their company’s data to create revenue, reduce costs and boost customer relationships.

“Although product, price, distribution or place, and promotion — the four P’s of marketing — clearly will always play an important role in business, they no longer create the advantage they once did — at least not a sustainable competitive advantage,” Nyberg said. “The ability to form relationships with customers and to nurture and grow those relationships is the ultimate competitive advantage — perhaps the only sustainable competitive advantage. If you have a fantastic product, your competitor can probably copy it, possibly within weeks — days if you’re in a service industry. But no competitor can copy the knowledge you have about your customers.”

Nyberg stressed the competitive value of service over product by quoting a recent study where only 8.3 per cent of those surveyed would not shop at a company again because of an inferior product, compared to 40 per cent who would not return because of poor service.

“I’m convinced that every industry is ‘commoditized,'” he told IT World Canada. “So how do you distinguish yourself from your competitor? With information and the ability to analyze your data to give better service levels than your competitor.”

For decision support, companies need database software that handles data, he added. He said a data mart — small-sized data that is tied to one application such as ERP, Siebel and PeopleSoft — can’t do more than 300 GB of data. Further, decision support problems arise when data comes from only one application, he said.

Seeing the customer’s view

He illustrated how many departments have some insight but no one has the big picture by depicting what might happen when a telecommunications customer repeatedly gets disconnected while using her cell phone during a conference call. Operations knows the call was dropped. Marketing knows this is a long-term, more profitable than average customer. Sales knows her contract is due for renewal within two weeks. Customer service knows from a sharp increase in customer complaints that there have been service issues in her calling area recently. Market research knows that a major competitor has just offered a new calling package bundled with a handset upgrade in an attempt to attract new customers and gain market share. Because of the segmented departmental data marts, the best decision not made quickly, the angry customer decides not to renew her contract and the company loses a long-standing, profitable customer.

“To build a relationship with a customer, you need to be able to respond with differentiated value at the point of need,” he argued. “You need to be able to turn the data you have about that customer – data your competitors do not have — into actionable information. If you can’t get the data you need, how do you even know if you’re making the right decision?”

Mark Hurd, NCR president and Teradata Division COO, joined Nyberg in attacking the use of data marts as opposed to enterprise-wide data warehousing. Hurd estimated that the total cost of ownership of a single data mart is $1.6 million to $2 million per year, including the cost of server, license fee or software maintenance fee for database, a couple of database administrators, plus network costs and facility costs.

He cited findings by New York -based Internet research firm BuzzBack that indicated the prevalence of data marts among large companies. Of the 108 IT professionals at a mix of industries but with 75 per cent in companies over $50 million, 70 per cent reported they did not have a single view of their data. Given that the majority of companies had many data marts — some even more than 100 – he calculated that 31 per cent of those interviewed had invested more than $20 million without getting a single view of business. He added that these investments add up rapidly because a data mart application “flies under the capital expenditures radar screen. It is done at the department level.”

Demand-based planning

At the retail level, insight into what the customer is doing fuels appropriate follow-up marketing — for example, sending patio furnishings promotion to someone who six months ago bought deck building materials, said Rick Schulz, vice president of Teradata’s retail, travel and transportation industry marketing. He pointed out that such knowledge can be used for store layout to get a higher number of items in the shopping basket as well as provide more convenience and service to customers.

Companies need to have the ability to recognize changes, respond and get back to suppliers within a couple of days, he told IT World Canada. He said that competitive retailers are doing merchandising planning by tracking the demand curve for a full year for a product so they can provide more accurate forecasting. They also need good mapping to the local market, he stressed, noting that different products have different patterns, and cognizant of local variances – for example, knowing how umbrellas sell in Vancouver and being aware of local festivals. “You can’t afford to keep the same level of stocking at all stores,” he said.

Schultz sees manufacturers facing the challenge of bringing data together from many business units and multiple countries to fend off global competition. He cites two trends. Not surprisingly, one is the use of an enterprise data warehouse bringing together procurement activities, product design and marketing and sales. He says companies are assessing design changes on quality of product in market. He gives the example of a manufacturer of a cell phone with 250 parts changing gone supplier of one part. “How does that change the serviceability in the marketplace after it is sold?” The manufacturer needs to know if the cell phone with that changed supplier ends up giving lots of service problems.

He says manufacturers need to tie together all points of data through to customer usage of product in order to provide customer-driven manufacturing that responds to service issues.

Schultz claims a second trend is better understanding of CRM among manufacturers, especially in the consumer packaged goods sector.

Schultz sees manufacturers and retailers building relationships of trust to share information, including real time point of sale data. As a result of this collaboration, a manufacturer may package items in a quantity of six rather than 12. Tracking sales through retailers allows CPG companies to be more customer-driven in terms of manufacturing.

Why do people buy?

People buy to solve a problem, says Richard Blackwell, Ohio State University marketing professor and author of the book Customers Rule!

He contends that a purchase always follows this process: problem recognition, search, alternative evaluation, choice of way to purchase, satisfaction/dissatisfaction.

“If a new product or technology does not solve a customer problem, it will ultimately fail,” he says.

Avoiding customer disappointment

In a report on “neighbourhood retailing” where retailers reproduce the personal service of a small local store, Teradata, a division of NCR Corporation in Dayton, Ohio, asserts that as retailers face their slowest Christmas holiday in a decade “the key to retailing success this year will be retaining the store’s customer base. Forecasting tools will be more important than ever in assuring that retailers stock the right assortment of goods in the right quantities for their customers.”

The company claims that by managing demand chain data – the record of your customers’ preferences – retailers can correlate purchases and best customers to better cater to their needs and preferences. Teradata also suggests sophisticated forecasting technology can catch fads earlier in the cycle, rationalizing the re-order process and allocating limited products to best customers.

According to Craig Petz, solutions consulting director of supply chain software developer SageTree Inc. in Irvine, Calif., supply chain management (SCM) should be based on analyzing current information, not just historical data so it is predictive rather than reactive. At the NCR Partners 2001 conference, he noted that this can be achieved by installing the maximum number of flow control points in an integrated SCM system to enable the live analysis and alerts that today’s massively parallel processing computing environments can handle.