BY DAVID GRABOVE

Many organizations are faced with the decision of whether to purchase off-the-shelf customer support software or develop custom applications internally. The build-or-buy decision is not unique to the support industry. It is a perennial issue faced by organizations in every business. Because of the relative newness of support applications, many are internally developed. Organizations have customarily fulfilled their own application requirements without giving much consideration to solutions available from other sources.

As information infrastructures become more standardized, the decision to build occurs less frequently. In fact, a Gartner Group study found that 50 percent of enterprises would only build if they could not buy, and 35 percent said that they would build only when it was strategically advantageous to do so. Extensive progress in integration and usability by pre-packaged application vendors has created a wide range of viable choices from which organizations can choose. Gartner Group's survey concluded that there is a dominant trend toward the purchase of software applications ("Trends in Building vs. Buying Applications," E. Keller. 30 July 1997).

What Do You Want to Do Today?
When deciding whether to build or buy a support center application, it is essential to determine application requirements and to assess your organization's ability to produce an effective solution, taking into account your available resources, as well as your core competence. Until recently, support centers were widely viewed as a necessary evil - a point at which to handle problems, complaints, and other everyday business tribulations. In that atmosphere, customer support software was needed to provide a reliable method for recording and documenting events.

In stark contrast, today's climate of downsizing and cost cutting is characterized by a burgeoning necessity for support centers to contribute to, and even increase company profitability. This fast-paced, competitive environment has spawned the need for customer interaction products with features that harness the newest technologies to eliminate unnecessary procedures and to streamline business activities. The requirements for modern support applications are far beyond those of simple logging and documenting:

• Tight integration with other systems and applications is required to provide Internet-based sales, technical support, and other information by linking remote customers, vendors, and users with a central database over the Web.
• Automatic escalation algorithms must integrate with e-mail, paging, and telecommunications systems to ensure that the right people do the right things at the right time.
• Computer telephony integration (CTI) helps eliminate redundant or unnecessary data entry and retrieval, while allowing efficient routing scenarios based on agent skill sets and availability.
• Current knowledge management technology learns, organizes, and retrieves information about past situations, eliminating the need to research the same problem more than once and giving support technicians a head start on problem resolution.
• Graphical metric displays and powerful report generators help support center managers better monitor and control their processes and apply strategic planning techniques to support services management.

Don't Overlook The Opportunity Costs
It is possible to build a basic, home grown call tracking system based on off-the-shelf database management software, such as MS Access, dBase, Lotus Notes, or HTML. However, these simplistic systems do not provide the functionality of sophisticated, commercial packages, nor do they incorporate the flexibility required to address changes driven by inevitable future growth and other environmental changes. Furthermore, the more comprehensive and complex home grown designs become; the more dedicated development resources are required; the greater chances are that the project will face delays or failure; and the more difficult future maintenance and adaptations will be. Some organizations are forced to build an application because they are retricted by legacy systems. As interoperability becomes the norm and more organizations adopt industry-standard databases and user applications, this limitation is becoming less common.

Even with adequate in-house programming resources, building a full-featured customer support solution may not be financially practical. According to the Gartner Group, a comparable problem management solution can take five full-time developers a year to build. Furthermore, such projects often end up costing more than planned because of unforeseen technology changes, changes in completion criteria during the project, or inability of the system to tolerate change during and after implementation (Gartner Group Research Note, "Strategic Planning," C. Lusher. 1 October 1996).

In addition to direct costs, the resources used to develop and maintain customer support applications withdraw from the organization's core effort.

On the other hand, when the same application is purchased from outside, both R&D and maintenance costs are distributed across the vendor's customer base, reducing each customer's application cost and increasing their return on resources. Using Gartner Group's time projection, five software engineers, each earning $58K per year could spend a year creating an application for their employer. For less than one-third of their salaries, you could purchase a 20-seat license and five years of maintenance from any of the top 38 support-software vendors¹. Some of those 38 products could be purchased for less than one-sixth the cost of developing a home-grown package (see Figure 1).

You could spend three times the cost to implement your system in one year; or you could implement it today at a fraction of the cost, while assuring yourself that the vendor will maintain and update the application regularly. While it is impossible to predict future maintenance releases, version upgrades, and platform migrations for a given application, it seems likely that for the cost of developing a full-featured support application internally, a comparable commercial package could undergo multiple complete design turnovers during a five-year period.

Time to Implementation
Implementing custom-built applications takes time beyond development cycles and involves more than the installation of software. To work in conjunction with existing databases or applications, linkages must be tested and proven. And, as with any new system, users need training. Finally, full implementation will inevitably necessitate some outwardly innocuous process changes that will later cause unforeseen problems.

Most commercial developers cite ease of implementation as a design objective. While some of their applications require vendor or programmer expertise to deploy, many do not. In a recent survey, 34 support software vendors indicated that it takes a typical customer an average of 20 days to make their application operational². Because customer support systems are strategically important, this time-to-market has significant impact.

Features and Functionality
A decade ago, there were virtually no software vendors offering customer management applications. The few applications that existed were either internally developed or custom-built to accommodate a specific mainframe or minicomputer system. As personal computers and network operating systems began to replace those more rigid platforms, the support milieu became more complex, and the help desk software business developed. Today, there are dozens of commercially-developed support solutions that range from basic, structured packages to complex, adaptable systems that require a programming staff to organize and maintain. To meet users' needs, some midrange choices incorporate the majority of features and industry best practices required, while affording the flexibility to customize the application to meet specific needs, without requiring programmer support.

Conclusion
There are now dozens of commercially developed support applications from which to choose. These applications offer feature sets and functionality ranging from simple, rigidly-structured, and easy to maintain systems, to elaborate, highly-adaptable packages that require considerable expertise to implement and maintain. Within that range, there are also products that offer a balance between customizability and ease-of-use, making those packages ideally suited for most customer support situations. Considering the abundance of high quality commercial packages, together with the costs and time required to develop, implement, and maintain a similar home grown application, it is often the most cost- and time-effective choice to purchase a customer support application from a vendor focused on producing and continually improving products that address the modern support environment's requirements.

David Grabove is a marketing analyst for Bendata, Inc., a developer of support management software solutions. Bendata's flagship product, HEAT, has an installed base of more than 3,500 customers and supports all aspects of help desk automation: call tracking, problem resolution, and problem management. Four principles guide the evolution of system design: ease of installation and use, total customizability, complementary add-on modules and integrations, and bringing enterprise concepts to a workgroup price/complexity level. Bendata's products are deployed across a broad range of industry segments with an estimated 50,000 end users at more than 3,500 customers worldwide, including industry leaders such as Alltel, Chevron, the FDIC, Mercedes Benz, and U-Haul. Bendata is based in Colorado Springs, Colo., and can be reached at (800) 776-7889 or at www.bendata.com.

Footnotes

1. Estimated average 20-user license is $48,112; estimated average maintenance cost is 20% of application price. Raw data taken from "1998 Guide to Problem Management Software," Support Management Magazine, May/June 1998.
2. Ibid.

Search the CTI, INTERNET TELEPHONY™, and C@LL CENTER Solutions™ Buyer's Guides for call center support management solutions.