When Southwestern/Great American Inc. set out to implement a customer relationship management (CRM) system from SAP AG, the Nashville, Tenn.-based company thought it had its ducks in a row. But when it came time to gather user input, a key business sponsor airily informed IT that having risen through the sales ranks years before, he would speak for the company's salespeople.
Project leader and senior business analyst Ryan Tabor was skeptical. "I told him, 'Even though you've done the job, we really should meet with the people who'll be using this [system] now,'" Tabor recalls. But the executive held his ground.
So when the SAP CRM went live at the $200 million-plus direct-marketing company, problems abounded. "We spent 75% longer in post-project support than we had budgeted, revising all these small processes," Tabor says. "The system had salespeople recording all kinds of activity stats that really didn't matter but reflected what a manager wanted to see."
Midmarket enterprises getting serious about CRM can learn plenty of lessons from larger enterprises, whose expensive, often highly publicized train wrecks caused analysts to estimate a few years back that only a fifth to a third of CRM projects would succeed. Moreover, midsized businesses earned battle scars of their own during their early CRM efforts, and they're now putting those lessons to use in an emerging body of best practices.
Done right, CRM is more than software; it's the practice of managing customers and data to increase customer satisfaction, retention and sales. Thus it requires senior management buy-in and efforts to develop work processes to best use the software to attain those goals. But it also requires an IT proponent who can manage aggressive business leaders to ensure that the end product reflects true user needs.
By the late 1990s, large companies were learning this the hard way. They failed to recognize that the toughest part of CRM was the cultural change it wrought. This led to confusion, unused applications and a sinking feeling among business managers that big IT is a waste of money. "Many large enterprises bought more seats and more modules than they needed," says Liz Herbert, an analyst at Cambridge, Mass.-based Forrester Research. "Some aren't even sure today what they have."
Today, large enterprises know CRM is more vital than ever; but many sense that they've spent years and millions of dollars spinning their wheels. "Even the largest companies want smaller deployments with less expensive pain," says Brian Prentice, an analyst at Gartner Inc. in Stamford, Conn.
That's music to midmarket ears. "The midmarket is more cost-conscious anyway, so there's a lot more focus on really rolling out what you need," Herbert says. "Midmarket companies are starting out small, maybe with [sales force automation], and getting it right before they move forward."
Such modest goals mark a significant shift in the way companies think about CRM. After all, the early promise of the technology was to link disparate groups -- so that, for example, marketing could put a full-court press on customers that the salesforce deemed close to a purchase decision. It's not that CRM pioneers were fools; it's more that real-world experience has shown that smaller, targeted projects are a more likely path to success.
This insight is key in a world where CRM software has taken off. In an August 2005 survey by Bain & Co., 75% of the 960 business and IT executives from both large and midmarket enterprises say their companies use CRM, up from 35% of respondents in the 2000 study. And from a list of 25 management tools, respondents ranked CRM second in usefulness, edged out only by the category of "strategic planning."
CRM's rising popularity is also reflected in the growth rates for its vendors -- particularly those that offer a hosted software model, which supports the move toward simpler and faster deployments.
Siebel Systems, whose hosted product was one of the few divisions performing well for the company in 2005, reported an increase of 34% in license revenue for the quarter ending Dec. 31 compared with the same period a year earlier. (This was the last quarter of independently reported revenue for Siebel, which has since been acquired by Oracle Corp.) Hosted application provider Salesforce.com grew 97% in 2004 and was poised for even higher growth in 2005. RightNow Technologies boomed 83% in 2004. And SAP released an on-demand product in February.
How to Buy a CRM System
| Midmarket companies need customer relationship management (CRM) systems with user-friendly features, so they should engage users in the requirements process.
But not too much.
In a typical scenario, so many users are polled "that you end up asking for a ton of features, which in turn tends to get you into a big vendor's big CRM package," says Brian Prentice of Gartner Inc.
And that's where the problem for midmarket IT comes in: Bigger is rarely agile, and agility is what gives the midmarket its edge over larger competitors.
Bigger does mean a pile of features that can sit idle. (Prentice says that only 28% of SAP CRM features get used.) It also means IT resources and a budget that can't keep up with the hefty app.
So rather than saying yes to every feature request, midmarket IT is better off going with a simpler application that it can get running quickly and add to later. Prentice says that Microsoft, Salesforce.com and FrontRange stand out in this area. "They're not obsessing on functional breadth and depth," he says. "They're building out implementation tools and [partnerships] for add-ons. They've realized they cannot beat the big guys at features and functions, so they're focusing on rapid modifications."
So how can you get started? First, ask future users to diagram their processes. Next, learn whether they're seeking to automate that process or whether they want to modify it. If the latter, work with users to diagram the new process.
Jim Benson, CTO at K/P Corp., says all this diagramming played a major role not only in his eventual decision to use Salesforce.com's hosted CRM software but also in the way he sold the process to business managers. "I did an elaborate diagram showing how many manual touches we had [without CRM], how salespeople's lives would get easier, all the info we could extract from our ERP system and so on," Benson says. "It showed we weren't just adding one more thing to maintain."
Paving the Way
"I'm not going to call our first [CRM] implementation a disaster, but it did not yield the productivity that it could have," says Nick Voutsakis, CIO at Glenmede Trust Co., a midmarket Philadelphia wealth management firm with $15 billion in assets under management for more than 1,400 clients. He hesitates, laughs and says, "I guess it wouldn't be wrong to call it a disaster."
Its first time around in 2001, Glenmede installed a package from a small vendor that "was bought out by Sybase, and then Sybase dropped the product," Voutsakis says. "So we supported it ourselves for two years." But, he adds, the application wasn't the true problem. "The CEO dictated, 'You are going to use this,'" Voutsakis says. Nonetheless, he doesn't blame business execs for the failure; put off by the CEO's edict and with insufficient knowledge about CRM, IT failed, Voutsakis admits, to help prospective users understand how CRM could help them better manage their accounts. It also didn't seek business units' input on functionality or the interface. The result: an unused CRM system.
Glenmede's experience reflects what often happens when CRM systems are imposed from above or adopted only because it seems like the thing to do, with little emphasis on securing executive sponsorship or user education. That's as true in midmarket companies as in larger organizations.
At North Shore Credit Union, the first CRM go-round in 2002 earned CIO Fred Cook an earful from tellers. The U.S. $56.5-million financial institution based in Vancouver, British Columbia, had a swank new CRM system from Pivotal Corp. The IT group failed to gather sufficient user input, Cook admits. As a result, it developed only a single data view of customers; tellers saw all the same CRM data that a loan officer or other senior exec would see.
North Shore's tellers immediately complained that this view overwhelmed them; with only a minute or so to complete transactions, they couldn't waste time paging through screens for information that was relevant to them. "They didn't have time to really look at all the CRM windows," Cook says. "They were looking at too many boxes [of customer data], and it wasn't really useful to them -- it didn't flow."
That wasn't Cook's only problem. Before he signed on at North Shore, business managers decided to encourage adoption of the CRM system by measuring the quantity of customer information employees entered in data fields. "We got garbage in, garbage out, with no substance," he says.
Pet Peeves Versus True Needs
Cook learned the hard way that less can be more. So when he led a redesign of the Pivotal system to give tellers a smaller menu of information, he also gave it to them in an instantly readable form and ensured that the data was useful to them.
The key to getting the system right, Cook says, was essentially to ask tellers to design the interface. "That screen went through at least four renditions," he says. "By the time it was done, we had real strong buy-in from the tellers. They saw this as their tool, not for managers, not for loan officers."
The focus on the end user surfaces repeatedly in discussions with successful CRM adopters. Jim Benson, CTO at K/P Corp., researched why CRM systems fail before he undertook CRM at the $75-million direct-mail company based in San Ramon, Calif. He and his team decided the key to success was rapid adoption by the salesforce and thus focused on streamlining the Salesforce.com interface. Training was exhaustive, and the company's vice president of sales served as the head cheerleader. The result, Benson says, was an 80% adoption rate within three weeks of the switchover.
"If you want to be successful, heavily weight [features and functions] toward salespeople in the field in your first phase," agrees Ken Morris, CIO at International Specialty Products, a $1.4-billion chemical company based in Wayne, N.J. "Management comes later."
Of course, Gartner's Prentice notes that too much user input can spiral out of control. Convinced of the need to pull end users into the project early, large enterprises have run into problems when user meetings devolve into pet-peeve sessions and the IT group is then loath to scratch off any requests. A ballooning list of requirements leads to a project scope with so many features that only big-foot applications from the largest vendors are considered.
Yet project leaders shouldn't be afraid to reality-check user requests. International Specialty Products' Morris has a team that evaluates user requests for modifications to the software on a monthly basis. "You want to see what's really a win versus what's just bugging one person," Morris says.
Southwestern/Great American's SAP system sat largely unused for nine months as sales reps refused to jump through the hoops created by a manager who claimed he understood their needs. For example, the manager hadn't thought to automate the task of sending marketing materials to prospects after a sales call. So in the first iteration of the project, this simple procedure required five steps in two separate applications.
Tabor's group got a second chance when the manager in question was replaced, and the new manager was willing to listen to IT's recommendations. This time, Tabor insisted on extensive needs gathering. "First, we held meetings to discuss processes that were new or being refined," he says. IT would then create mocked-up screens. Then it would leave those screens in the meeting room with end users and no IT staffer present. "It's one thing to stand up there and explain, and another to leave them alone to interpret the [user interface]," Tabor says. "If the way they think it's going to work is the way it's actually going to work, you've done it right." If not, it's back to the drawing board.
CIOs agree that the day you stop training and cheerleading for your CRM system (and tweaking the software itself) is the day it starts to wither -- even if it's been up and running for years. Enterprises with effective CRM issue new releases every week or so when the system is first rolled out; eventually, these releases taper off to every six weeks.
Whether they've been knocked down themselves by CRM projects or have merely read about others' problems, midmarket CIOs are demonstrating that they can internalize lessons and move forward. "CRM is a cultural tool, not a product tool," says North Shore's Cook. "Once you figure that out and proceed accordingly, everything else falls into place."
For more on CRM, see "IT Management Guide for SMBs: CRM".
Steve Ulfelder was a senior features writer at CIO Decisions. To comment on this story, email firstname.lastname@example.org.