Upgrades on the Upswing

It’s no secret that many CRM rollouts are dismal failures – or worse. If you’re wondering what could be worse than a dismal failure, try this on for size. Take a CRM rollout that didn’t catch fire; now upgrade the program, and you have double dismal. Sort of like double-digit inflation. A geometric progression of dismal proportion – technology-wise.

A few months ago Gartner Group, the world’s largest high-tech market research firm, made a startling pronouncement. According to their analysis of past trends and current directions in the area of CRM, “through 2006, more than 50 percent of all CRM implementations will be viewed as failures from a customer’s point of view.” While Gartner analysts have been known to be wrong in the past (these are the same guys who predicted the world would collapse on Y2K), such a dire prediction bodes ill for companies depending on CRM to pull them out of the economic doldrums.

Gartner’s warning is all the more poignant because CRM is undergoing a fundamental shift in the way systems are implemented. Most of today’s CRM systems are based upon what’s called a “client/server architecture,” which means that there is a really big database program running in the data center (the “server” computer) and a smaller user interface program running on the salesperson’s laptop or desktop (the “client” computer).

When client/server was originally introduced, it was a vast improvement over the now-ancient mainframe-based CRM system that ran on character-only terminal screens. Client/server, however, turned out to be a major headache, especially when it came to upgrading the software. Because the client portion of the CRM system was actually installed on every salesperson’s computer, making changes could involve installing software on hundreds or thousands of individual salespeople’s computers.

Most of the new CRM systems reject the old client/server architecture in favor of Web-based programming, which uses the same technology as the Internet. In addition to speeding the development process, Web-based technology makes upgrading a CRM system far easier because the salesperson’s computer only has to have a single piece of software mounted on it – a Web browser. So a Web-based CRM system can be upgraded with new features simply by loading new pages into the Web server in the data center.

The good news is that these Web-based systems, when installed, will remove a lot of the pain of upgrading. The bad news is that upgrading from an old client/server system to a new Web-based system could become a major headache to companies that haven’t thought through the process. With this kind of major transition on the horizon, sales groups need to make some key decisions in order to be certain that any future CRM upgrades will help their company be numbered among the successes in CRM.

Key Decision #1: Upgrade the Existing System or Start over from Scratch?

Any kind of major CRM system upgrade is an opportunity for a sales group to reassess whether it has the right CRM system for its needs. There are two key dimensions to this decision. The first is whether the current system – after it’s been upgraded – has the features and functions that the sales force needs to become more productive. Obviously, if the upgraded system isn’t up to snuff, then it makes sense to consider another vendor’s technology. “When moving to a new upgrade, you should expect to have enhanced ease of use, power and functionality,” says Cheryl McCarthy, executive vice president of marketing at Surado Solutions Inc., a CRM vendor in Riverside, CA.

The second dimension is the total cost of upgrading. If the cost of upgrading begins to approach the cost of installing a new system, then it opens the door for newer technology. “For systems that are owned in-house, upgrades tend to be about 20 to 50 percent of the original cost to install the software,” says Erin Kinikin, vice president at Giga Information Group, a firm that consults for customers of high-tech products and services. Installing brand-new software is likely to be more expensive, but may be worth the extra expense if the new system is better suited to the sales group’s needs.

It’s also important to look not just at the basic cost of the software but at the total cost of ownership (TCO), according to Laurie Orlov, research director for business applications, at Forrester Research Inc. “Companies need to think about the cost of training, the cost of connecting CRM system in the event of a merger or acquisition, and the cost to convert data from the old CRM system to the new,” she explains.

Key Decision #2: Own or Rent the System?

It used to be that the only way to get at a CRM system was to buy the hardware and the software and install it in a data center staffed by the company’s own employees. That’s no longer the case. CRM vendors offer a variety of business arrangements that range from the old-style buy-it-and-own-it model to the Application Service Provider (ASP) model, where the sales group just accesses and updates sales data stored on the vendor’s servers on the Internet. Other arrangements include outsourcing, where the vendor (or a third party) purchases and maintains the CRM system, and software by subscription, where the sales group pays a monthly rental fee for software that’s installed on in-house hardware.

Sales groups need to decide what pricing scheme is right for them. With a rental agreement the up-front cost is less, but there’s always the possibility that the sales group could, over time, end up spending much more on software than if it had been purchased outright. “It’s very much like making a decision about owning or leasing a car,” says Jim Garden, vice president at Technology Business Research. “If you decide to own, you pay a greater up-front cost, but the car belongs to you.”

Indeed, this may be a particularly good time for sales groups to pull their vendors to the bargaining table. A recent survey by market researchers Dataquest Inc. indicated that worldwide CRM software revenues declined 8 percent in 2001 compared to revenues in 2000. With CRM vendors struggling to weather the economic downturn, it’s not surprising that many vendors are willing to make a deal. “Sales groups shouldn’t be afraid to negotiate for the kind of deal that makes sense for their own cash flow,” says Garden.

Key Decision #3: Redesign the Processes or Customize the System?

Many companies purchase a CRM system in order to speed up their current sales processes. Because those processes may have accumulated over a number of years, creating an exact match in the CRM system generally requires a great deal of customization. While reluctant salespeople may be more willing to use a CRM system that emulates existing sales practices, such a conservative approach may limit the gains in productivity that might otherwise be possible.

For example, if a company’s pre-CRM sales process required the salesperson to check with a warehouse supervisor to determine product availability, a CRM system could easily be customized so that the supervisor continued to be the bottleneck for processing sales. However, it would probably be more productive to have the CRM system directly access the warehouse database so that salespeople could immediately determine whether the product could be shipped. While that may be an overly simplistic example, there are many cases where companies have used CRM to reinforce procedures that are decades old, simply because that’s the way that things have always been done.

Most companies overestimate their need for customization, according to Keith Raffel, CEO of UpShot. “We’ve talked to some companies whose sales and support processes are so involved that we’ve told them they need to go higher end and get something custom made,” he says. However, such companies, in Raffel’s view, are “few and far between.” He feels that most companies can probably be better served through minor system customizations combined with judicious changes to the sales process.

Flexibility is a key issue, according to Surado’s McCarthy. “Companies need to look at their corporate culture and assess their ability to manage change,” she explains, adding that even relatively conservative companies can redesign their processes to be more efficient if they get key sales personnel involved in the decisions. “It can’t just be the CEO and the sales management forcing a change down the chain of command,” she insists. “You have to get the entire chain to buy into the change, because the chain is only as strong as its weakest link.”

Key Decision #4: Full Testing or Take a Big Risk?

One of the biggest mistakes that companies make when upgrading is not testing the system sufficiently, according to Steven Sherkanowski, director of product management at Pragmatech Software Inc. “Ideally, upgrades should be staged in a test environment that matches a company’s production environment,” he says. “That way the sales group can assess changes to existing functionality and new functionality.” He notes that this kind of test-bed environment is generally practical only when companies own their own CRM systems. However, he points out that major upgrades of ASP-style CRM systems could be brought up on a separate Website for testing purposes, allowing sales groups to assess new features.

Preproduction testing also allows the company to assess the amount of sales training that will be required to get the salespeople productive on the new system. However, companies should look askance at any new system or upgrade that requires more than a few hours of sales training. “CRM is an optional-use application because nobody is going to fire top salespeople just because they aren’t doing their electronic paperwork,” says Kinikin. She believes that it’s critical that the salespeople actually enjoy using the system. “Sales and marketing people don’t have the desire, or the patience, to become programmers,” she says. “The system has to be intuitive enough so that they can use it even without a lot of classroom instruction.”

It’s also important for companies to do load testing in order to determine if the upgraded system will run efficiently in a production environment. (In a load test, the CRM system is artificially loaded with users and transaction to see at what point it begins to fail.) Such testing is particularly important if the upgrade involves converting from a client/server architecture to a Web-based implementation, according to Kinikin. It turns out that Web-based system browsers aren’t as efficient at updating screens as the old client/server applications. “Customer support can cost from $50 to $100 an hour,” she says, “Nobody wants a system that forces people to wait 10 seconds to repaint a screen when a single item of data changes.”

It remains to be seen whether Gartner’s dire prediction about the future success of CRM will actually come true. However, it’s clear that companies who manage to upgrade and grow their CRM capabilities are more likely to be counted among the successes rather than the failures.