Plenty of corporations have plunked down a big chunk of change on snazzy customer relationship management systems to milk current customers for all their worth and capture new ones. Unfortunately, all too many of them have woken up the morning after with a huge migraine of a systems integration problem.
The explanation is simple enough: When new software installations aren’t properly thought out in advance, they rarely work as advertised. Meta Group, a Stamford, Conn., based market research firm, recently issued a report that outlined some of the things that can go wrong with a CRM system if it is not planned correctly or doesn’t meet a company’s business goals.
Meta Group is hardly alone. Analysts at other research firms say they, too, have seen companies stumble their way through a CRM project.
“In a perfect world everything would work together from the moment a company contacts a customer with a promotion through to when the customer makes the purchase, receives the product and gets support,” says Erin Kinikin, a vice president at Giga Information Group. The problem is that delivering on that promise is a five to ten year IT project.
Getting front and back-office applications to work in harmony is an even more monumental task, Kinikin comments, and it is just as much a business-reengineering task as it is a technology challenge. Too many companies have been caught up in the bad habit of merely adding software to automate their existing processes without reengineering them.
“You can have all the integrated data in the world, but if employees are not trained to ask ‘what can I do to make an at risk customer come back to us’, then you’ve spent millions of dollars for nothing,” Kinikin stresses.
If the business processes that underlie a CRM system are fragmented, then the CRM system will never work no matter how well it’s designed.
But collaboration across departments at the process level is inhibited by differing incentives, Kinikin contends. Sales, marketing, and customer support staff incentives are based on different criteria.
For example, a sales person’s commission might be based on the volume of sales they make. A marketing person might get a bonus based on the number of leads they generate.
“Marketing is [thus] motivated to give sales leads that aren’t good, sales is motivated to sell deals that may never be profitable and customer service is motivated to get off the phone quickly instead of asking the one cross-sale question that would result in a more valuable customer,” she says.
Gareth Herschel, an analyst with Gartner Inc., a market research firm that is also based in Stamford, says, “If sales and marketing are not communicating with one another, then there is no way sales applications will talk to marketing applications.”
So how can companies get out of the mess of being stuck with a hodge- podge of systems?
Giga’s Kinikin recommends that businesses should let the customer drive the degree to which CRM applications are integrated with the back office. It seems ironic, but businesses do not take enough of a customer-centric approach to CRM.
“If there needs to be integration with the back office because customers are dying to know whether you’ve shipped the product they ordered, for example, then selective integration makes sense.”
Gartner’s Herschel says many companies have chosen a best-of-breed approach to CRM because no one software company has yet developed a package that is strong across the entire customer life cycle, and that has limited the extent to which corporate users can integrate their CRM systems with other applications.
Although some software companies market their CRM products as “integrated suites,” says Herschel, they are strong in some areas of CRM, but inevitably weaker in others.
“Oracle, for instance, is strong on the sales and services side but weak on the marketing side,” Herschel says. “E.piphany is much stronger on the marketing side but weak on the sales and services side.” So companies pick and choose, and end up with a mélange of disparate applications.
Companies can integrate the systems themselves, but that can be very complicated and expensive, says Aaron Zornes, research director at Meta Group. An in-house development staff doesn’t come cheap, and application vendors aren’t always helpful providing the tools systems integrators need.
On the other hand, a company can purchase an integration package from a leading vendor such as CrossWorlds, Vitria, WebMethods, or another firm.
“The real benefit [integration packages] bring is speed to integration, particularly of stand-alone best-of-breed applications,” says Gartner’s Herschel. “They make the best-of-breed approach a lot more viable for enterprises,” he adds.
But Zornes isn’t so sure. He claims that although it is conceptually appealing, it can be very expensive, and does not provide full integration across all applications.
“You basically have to pay a quarter million dollars for an adaptor, that’s the ballpark market. You then get some business level connections, but you don’t get full integration between the packages,” he cautions.
Finally integrated suites are just not as integrated as the ERP vendors insist. In practice, they require a significant level of customization and integration with legacy systems. In addition, because applications that have been canned into suites are generally not all strong across the board, they require companies to “settle” in the areas where the suite is not as strong as a best-of-breed application would be.
At the end of the day, the route a company takes depends upon where its priorities lie, says Zornes.
One thing is certain. The integration headache will not go away overnight.
![](https://pixel.welcomesoftware.com/px.gif?key=YXJ0aWNsZT1iMDZlZWIwNDJmYTExMWVjYjIyNzJhZDE5YjQyODAzYQ==)