A CRM purchase should always be preceded by a needs-assessment analysis. These are complex endeavors that map out a company’s pain points in its customer-service operations to determine what application or series of applications can address these problems.
After a few years of declining sales, companies are beginning to step up purchases of customer relationship management (CRM) applications, adding new modules to preexisting applications, shifting over to the hosted method or, in the case of a newly formed company, purchasing a basic suite for the first time.
The choice for many of these firms will be Salesforce.com, a company that leads the hosted CRM market, according to numerous surveys, and SAP and Siebel are close behind.
But no matter what vendor these companies select, those responsible for purchasing the software need to heed a few cost basics to ensure their new CRM projects don’t go off track. The Forrester report noted that companies will spend a total of US$13 billion on CRM initiatives in 2005, but only 25 percent of that will go toward software licensing, with the balance going toward integration, administration and maintenance.
If these numbers sound familiar, that is because they are. Almost all enterprise software projects break down along these lines, with only a small portion of the overall expenditure devoted to license fees. Of course, this is partly due to the integration and administration associated with rolling out any new system. But in some projects, these categories often turn into financial black holes.
Before You Buy
A CRM purchase should always be preceded by a needs-assessment analysis. These are complex endeavors that map out a company’s pain points in its customer-service operations to determine what application or series of applications can address these problems.
Analyzing historical customer interaction and transaction data is part of that process. Such a step will ensure — at least theoretically — that the application purchased does not require more training than expected or will overwhelm the company’s infrastructure.
Unfortunately, companies do not always do a great job of evaluating the capabilities and limitations of their existing hardware and infrastructure when planning an enterprise-software purchase.
As You Buy
Earlier this month, Oracle announced it was changing its pricing methodology for hardware containing multicore processors, a shift that some have speculated is a response to adjustments in Microsoft’s and IBM’s pricing.
Pricing for software and hardware is quite complex, which means it can be easy for companies to leave money on the table when negotiating a purchase. It is important for corporate negotiators to be current on industry pricing trends, including what competing firms are offering.
Also, while most companies are savvy enough to negotiate discounts, they often are caught unawares by tricky maintenance and upgrades clauses.
After You Buy
It is important to assume that additional money will have to be spent on training. A great number of CRM projects have failed to deliver return on investment because a company relied on the vendor to deliver the necessary training. While vendor training is important, it should be only a part of a comprehensive training program developed by the buying company.
Companies tend to neglect training — especially ongoing training. One common metric — although cases can vary according to many factors, such as the work-force size and the complexity of the system — is that training should comprise at least 10 percent of the implementation budget.
The implementation budget itself, which covers the cost of integrating the new application with existing infrastructure, will be considerable. Another rule of thumb is that for every dollar spent on a software license, at least three should be spent on implementation and integration.
The final area to watch out for is extensive customization, which often leads to glitches and, indeed, further systems integration work.

















David … you make an important point here that my partners over at Accenture and I align with. CRM is a great, enabling technology BUT it is just the technology piece. The technology in and of itself provides little value. The real value lies in how it’s implemented and rolled into to an overall strategy and execution that includes people and process.
Frankly the technology is the least value creator piece of the puzzle. That’s why the services piece will always be the greater cost portion … more value is derived (if done properly) from services than anywhere. I realize that’s a big IF.