Talk Is Cheap

Will customer analytics save SRM from becoming shelfware?

News Item: As its first commercial product, AT&T Labs introduces Natural Voices Text-to-Speech. Sporting a library of voices, the product can turn text into “natural-sounding” speech. AT&T calls the product a “fast and effective way to extend corporate image, brand, and personality through ‘made-to-order’ voices”. As The New York Times noted, the technology could let Chicago Cubs fans listen to the late Harry Caray call baseball games and serenade them with “Take Me Out to the Ballgame” into eternity. Perhaps soon, deceased celebrities’ voices will be auctioned off for as much as Van Gogh paintings, except that they’ll come with a lot more marketing potential.

Meanwhile, the U.S. Congress continues to engage in heated debate over what kind of legal and political restraints to put on human (in whole or in part) cloning. And on the Web, users are sprouting multiple IDs faster than they can sign up for new credit cards. As if the CRM industry didn’t have enough on its plate: In the future, the effort to gain a 360-degree view of the customer may become a cacophony-riddled exercise in schizophrenia. For we can be sure that customers will also gain the ability to respond in kind to automated phone pitches done in the voices of celebrities: You call me during the dinner hour, and I’ll sic Bela Lugosi on you.

Judging by the headlines in the trade press, CRM solutions providers are indeed hurting, even without having to consider an exotic future full of multiple personalities, repeatable genetics, and thousands of customers who prefer to sound like Greta Garbo. Meta Group and other market research firms are reporting that well over half of all CRM projects fall short of their objectives. Once bitten by a computer-generation mistake, many marketing managers have become twice shy about exposing customers to CRM follies. Other news stories report that in organizations forged by mergers and acquisitions, line of business managers are proving resistant to enterprise solutions that demand that they share vital customer data with potentially rival divisions and sales groups. For these and other reasons, CRMs in danger of becoming synonymous with shelfware.


As with most software applications, the immediate promise for return on investment (ROI) is efficiency. Increased marketing automation and customer self service can, for example, reduce 15,000 calls a day to support personnel to under 10,000, leading to significant saving. But too often, efficiency brought to one operation only moves the inefficiencies somewhere else, creating snags and bottlenecks in operations, enterprise software applications, and database resources that weren’t anticipated when these systems were set up.

An intelligent enterprise must consider the ripple effects of efficiency created in one area upon people, applications, sales channels, and customer relationships. Unfortunately, C-level business executives – the prime target of most IT vendors these days – too often reach the impact on everything else. Today, no new application is an island; CRM almost by definition involves numerous parts of an organization, from front office to fulfillment and logistics. Personalization, for example, can be great for customers and the organization: But it also carries the risk of confusing customers, miring them in bugs, bad data, and snarled business processes that send them quickly to competitors.

To understand the value of efficiency, companies have to be able to measure it. And to manage efficiency’s current or potential effects on customers and processes, companies desperately need metrics and analysis methods. However, the Web-infused world is unlike most companies’ experience with offline transaction processing systems. Standard reporting and database query solutions have trouble supporting a world that is withering under waves of terabyte-sized clickstream data files, which in some cases are arriving daily.


Promising an ability to analyze data in ways outside the logical constraints of database systems, data mining has found a happy home in the world of customer analytics. We saw a mini-explosion of companies that rose quickly to meet the pressing data analytics needs of dot-com organizations – needs that leading business intelligence (BI) solutions providers, weaned on traditional database reporting, could not meet effectively. Because many of the most interesting dot-coms built their business plans on the idea that they would gain an increasingly sophisticated understanding of their customers and markets through clickstream analysis, they invested heavily in data mining tools and services, which made the Web analytics vendors look like geniuses. Data mining algorithms love huge data files, from which business managers desire to know the intimate shopping behavior of heir online customer.

But now, with the tide turning toward click-and-mortar companies, Web analytics vendors are trying to avoid being left high and dry in a Web-only world. These vendors must adjust their offerings to support multichannel analysis, meanwhile keeping an eye on traditional BI players that are starting to field competitive Web analysis solutions. They must also adjust their prices for a broader market. DigiMine, for example, boasts a hosted services model to shield companies from the necessity of hiring a fleet of highly trained statisticians to interpret data mining results.

Not all clickstream analysis and Web analytics vendors are going to survive those tough economic times – another factor that has given business customers pause, thereby slowing adoption. No one wants to invest in data ghost towns or be forced to migrate extensive data marts from tool to tool. However, companies can’t effectively proceed with CRM until they have the ability to analyze what these applications are doing to customer retention, buying patterns, sales performance, and overall customer and partner relationships.

Businesses are starting to listen to their IT architects, who would like to acquire layered, flexible, multichannel customer analysis solutions: that is, solutions that fit the strategic direction for most enterprise applications. In future issues, I will discuss how some leading vendors are meeting these challenges. And I promise I won’t do it in the voice of Bela Lugosi.

By: David Stodder

Source: Intelligent Enterprise


Leave a Reply