Companies are tiring of CRM projects, which are failing at a rate of 70 percent, but there is hope. Jill Dych\u017d, author of The CRM Handbook: A Business Guide to Customer Relationship Management (Addison-Wesley, 2001), says CRM would work better if companies did a better job defining specific needs and customizing the technology to fit them. She spoke with CIO about CRM\u2019s problems and potential. (For the full text of this interview, go to www.cio.com\/printlinks.) \n\n \n\n\n\n\nCIO: Are we about to see a CRM backlash? \n \n\nDych\u017d: It\u2019s already started. Companies are canceling CRM projects because budgets have gotten way out of whack, and they\u2019re renaming projects, shying away from the CRM acronym. They call it their corporate customer loyalty program, customer reengineering program, customer care initiative or even just campaign management. As these programs succeed, we\u2019ll see a readoption of the CRM term and an understanding of what it really means?a combination of process refinement, new information and organizational change. \n\n \n\n\n\n\nWhat mistakes do companies make with CRM implementations?\n \n\nThe biggest mistake is to let technology features drive CRM functionality. Instead of defining the business problem up front then locating the technology to solve the problem, these companies are doing it backward. Another problem is thinking that because you have a customer database, you are doing CRM. Data warehousing and decision support are not equivalent to customer relationship management. Likewise, just reorganizing around a new customer focus doesn\u2019t mean you\u2019re doing CRM. CRM means changing your focus and the business processes to support it, and applying technologies to automate those new processes. By nature, CRM is customized. There\u2019s no such thing as off-the-shelf CRM. That\u2019s a controversial statement, but it\u2019s true. \n\n \n\n\n\n\nIf companies already have off-the-shelf CRM, how can they make it work? \n \n\nBack up and redefine the business problems, the ideal business processes and information to solve that problem, then automate the tool around the information and the business processes. \n\n \n\n\n\n\nAnd what if a company is just starting a CRM project? \n \n\nDefine the problem, understand what\u2019s going to solve it and understand the functionality of the solution. Once you understand a requirement, you can find technologies to support that particular functionality. That\u2019s the right way to do it: the requirement, the functionality, then the tool.