The Truth About CRM Success & Failure
When Carl Underhill accepted the job as chief information officer of Pioneer Investments in April 2001, he began his first day staring failure in the face. The global financial services firm that hired Underhill built its reputation on making choice investments--that is how the 74-year-old company grew to manage $98 billion in assets, employing a 90-person sales force selling to 40,000 financial advisors. However, even Pioneer's successes in financial investments could not prepare the company for what was to come with its CRM investment.
In early 2000 Pioneer decided to proactively manage its client contacts, sales and marketing information, and new-product testing results by implementing a CRM solution. Pioneer's goal was to provide rich data about client behavior and competitors to internal marketing and sales staff, as well as to field sales teams, via a Web-based solution.
After examining various products, Pioneer bought a full CRM footprint from E.piphany Inc. The decision was based largely on E.piphany's strength in Web-based services, which was important given Pioneer's large field staff. Additionally, E.piphany wanted to build its financial services clientele, so it was amenable to helping Pioneer mold and tailor the product specifically for its needs.
Pioneer began to implement the E.piphany suite in July 2000, but within months it became clear that the solution was too big to make for a successful, timely, on-budget rollout. The plan stalled by November and sat idle until Underhill's arrival six months later.
With the findings of well-respected research firms, such as Gartner Inc. and Meta Group Inc., which state that anywhere from 50 percent to 80 percent of CRM implementations fail, many would likely think Underhill's chances for a successful CRM implementation were slim. These statistics have thoroughly scared top-level executives to the point of either abandoning existing CRM implementations or halting plans to begin one. However, CRM consultants are more optimistic about CRM success rates, as they claim inconsistencies exist not in the analysts' research, but in their definition of failure.
Despite CRM consultants' efforts to allay their customers concerns, the research firms' statistics quickly spread throughout the industry. This sent shock waves that are still reverberating today. "Companies are looking at those numbers and saying, 'never mind' to CRM," says Sue Handman, CRM solutions director at RCG Information Technology, a consulting firm based in Edison, N.J.
Barton Goldenberg, president of ISM Inc., a Bethesda, M.D.--based CRM consultancy, has performed more than 300 CRM implementations with a success rating of more than 90 percent. Even so, he is also finding more reluctance toward CRM implementations from his clients: "I have customers asking me, 'With the failure rate at 80 percent, why do you think this project should succeed?'"
While top-level executives are swayed by the broad failure statistics, confident consultants are not. "I vehemently disagree with the statistics. Not only are they unfair, but they are doing unreasonable damage to the CRM industry. These numbers are damaging top-management perception levels," Goldenberg says. "I'm not knocking the leading IT analysts. It is their business to provide general reports of interest to their paying client-base. What concerns me is the way they measure those success rates."
Handman concurs: "I agree that the analysts who reported these statistics believe them, but I don't believe they are right. I believe that defining failure is the issue."
Explaining the Numbers
Last fall Gartner issued a statement that said, "Through 2006 more than 50 percent of all CRM implementations will be viewed as failures from a customer's point of view." In this case the customers are the companies that "are receiving the benefits of these processes," says Scott Nelson, a vice president and research director at Gartner.
Despite common misconceptions that 80 percent of all CRM implementations fail, Nelson says that the 80 percent failure rating only applies to sales automation implementations. "CRM should provide salespeople with better pipeline reporting, rather than only make it easier to sell more. The latest CRM solutions are forcing salespeople to enter more administrative numbers than before. As a result firms find they spend millions in sales automation only to learn that sales reps are still using ACT," Nelson says.
Yet, not all components of CRM have such a high failure rate. "It depends on what subset of CRM you are implementing. Email response management systems tend not to have that kind of failure rate. The 50 percent failure rate was across the spectrum," Nelson says. "We wouldn't label it a failure unless one aspect was so overwhelming, such as a drop in customer satisfaction, shopping cart abandonment rates go up, or sales decrease."
Nelson concedes that measuring CRM failure is often difficult. The most common problem is many companies implementing CRM "don't apply good project management skills" and do not set metrics that can help them define success or failure, Nelson says.
This is where consultants find holes in the research firms' assertions. Without a predefined set of mutually agreed upon metrics for a CRM implementation, failure cannot be accurately measured, consultants maintain. Without these metrics failure rates are subjective, because they are based on what matters most to the individual people involved and not on a mutually agreed upon set of values.
To prove how the same CRM implementation can be viewed as both a failure and a success, Goldenberg cites one client that is implementing a sales and marketing CRM project involving more than 3,000 people. "When asked if it was successful, the marketing people said for marketing purposes the global customer profiles were a raving success. The CIO, however, said the CRM project was a dismal failure, because eighty-four databases were not properly integrated into one holistic data structure," Goldenberg says.
Nelson acknowledges the subjectivity of Gartner's findings. "There's a certain element of subjectivity, because there were multiple analysts participating in different areas, and as a result there isn't a unified definition of failure. This was not one monolithic quantitative study. It was pieces of studies put together," he says.
So the question remains, How does one define failure when implementing a CRM solution? "The failure rate is driven by not clearly understanding what you want to accomplish, and then designing your project to focus on, and deliver, results that are not aligned with your goals," says Ray Hicks, senior consultant at Kowal Associates Inc., based in Boston.
According to RCG's Handman, to define failure it is essential to first define success by outlining expectations before implementing the CRM solution. "Failure is the inability to meet mutually defined expectations. A lot of customers haven't set these expectations. Everyone, including IT, salespeople, customers, the software vendor, and integrators, has to agree on what the expectations are, how the project will be implemented, and over what time frame. And if they can't define the end product they should define it in pieces," Handman says.
Breaking up the CRM project into smaller, more manageable components is critical, because business needs change over time. But not all customers recognize this initially. Goldenberg recalls one customer, Jim Rubin, CIO of Tripos Inc., a provider of drug-discovery chemistry, software, and enterprise informatics, who initially refused to set a business case with metrics prior to Tripos' CRM project. "Nearly every company I've been in is not static. You need a business plan that takes that into consideration. If we change our business, you can't track that against your business plan, because you're comparing apples to oranges," Rubin says. "I pushed back pretty hard on that. I'm not particularly a fan of a report that sits on a shelf and doesn't get used."
"He was heading for a shipwreck," Goldenberg says. However, Goldenberg persisted, and with Handman's needs in mind, he suggested writing a business case with smaller, more attainable goals over a shorter period of time.
Goldenberg's persistence paid off. "I can't claim that I fully respected Bart's approach when we retained him; however, [implementing CRM] in stages and demonstrating the benefits to each stage is critical to the success," Rubin concedes.
Trying to implement the whole E.piphany suite at once, instead of in manageable pieces, is what set Pioneer off course, Underhill asserts. "When I came in, we took a look at the initial plan and made a 180 degree turn," he says. "Besides being too large, the major issue was that we couldn't get the [CRM] project map on a white board and draw a business benefit against every feature and function the suite offered. It didn't make sense to implement everything like we had been trying to do."
Creating a Plan for Success
Underhill's plan was to isolate the portions of the E.piphany suite that could provide true and immediate business benefit and start with those. "The technology wasn't an issue since E.piphany lends itself to scalability. The biggest issue was simply trying to be all things to all people," he says. "We took a small project team and spread everyone so thin that no one could rapidly address changing business needs. Breaking the implementation into manageable parts solved that." But it was a task the CIO found easier said than done.
"Rolling out portions of CRM solutions is difficult--they are so integrated and built around the idea of end-to-end. To cut them up, your approach has to be careful," Underhill says.
That careful approach began with creating workgroups of sales and marketing teams, senior management, and the IT staff to hash out which CRM functions had to be available immediately and which could wait.
The group first determined sales force analytics was the most immediate need. The goal was to take contact management in its purest form and add features of CRM so the field sales staff could get rich information to aid the sale, with a delivery method that would replace telephony and hard-wired laptops.
To achieve this Pioneer invested $15,000 in Blackberry pagers last September and developed software to integrate portions of the E.piphany sales analytics solution with the pagers. Now salespeople can query the system for information about a client, such as its last big purchase. There is also an alert system to push to users information that they specify themselves.
Underhill says that now he is confident that every dollar spent on the new technology will be earned back. "Even if building a CRM solution in this manner doesn't allow us to save money, it allows us to use our funds efficiently," Underhill says. "We are doing a smarter campaign that will translate to higher sales later."
Despite some of the damage that CRM consultants claim the research firms' reports may have done to the industry, Gartner's Nelson argues that the reports have helped the industry. "We have some people accuse us on the hype cycle that it pushes CRM into the trough," Nelson says. "However, it makes people realize CRM has to be viewed as more than just a technology. It helps firms become more conscious to properly staffing projects and demanding more rigorous ROI work. It has helped the market mature." Nelson says that he has seen this firsthand: Clients are more informed about CRM and are
asking him better questions.
Pioneer and Tripos are prime examples of how organizations are getting smarter about implementing CRM, and more success stories like theirs are cropping up. In fact, analysts at the Aberdeen Group Inc. are determined to point that out in a report called, "What Works: Ten Successful Implementations in CRM/Customer Contact Centers," which is slated for release this summer.
"We expect to find that CRM really works," says Christopher Fletcher, vice president and research director at Aberdeen Group. "This is partly in response to the horror stories about the CRM failure rate. We believe there are a lot of positive benefits that come out of CRM implementations, both qualitative and quantitative."