Computers in Libraries [June 2006]

Comprehensive cost planning yields successful tech projects

Breeding, Marshall.

Copyright (c) 2006 Information Today

Image for Comprehensive cost planning yields successful tech projects It is important to consider all the cost components of a technology project. At least some of the costs may be hidden, and itís easy to overlook them. Trying to anticipate all the cost implications of a technology-related project is common sense. The total cost of ownership model can be applied to library technology projects. TCO just reminds us to think broadly and long-term. You donít have to be an economist to incorporate this broader view into the budget projections of a library technology project proposal.

No library that I'm aware of has money to burn. Though funding levels remain modest, libraries continually have demands to improve existing services and to deliver new ones. Increasingly, these services involve the use of technology. Consistent with the theme of this particular issue, Technology on a Budget, librarians must often find ways to implement technology projects with very limited budgets.

Given these economic pressures, it's important to consider all the cost components of a technology project. At least some of the costs may be hidden, and it's easy to overlook them.

I've been involved in a lot of technology projects over the course of my career at Vanderbilt and through the numerous organizations that have engaged me as a consultant. From all of these experiences, I have gained some perspective on what's involved in trying to accomplish important work but having limited resources to carry it out. Most of all, I've learned it's far better to recognize the full impact of taking on a project upfront rather than have to deal with the ramifications of unforeseen expenses after the effort is already underway.

While most of the points that I offer may be obvious, I think it's important to lay out a complete picture of the direct and indirect costs of a tech project.

Total Cost of Ownership

One of the catchphrases in the many information technology trade journals that I scan every week is the "total cost of ownership" of adopting a new technology. These publications advise buyers to consider all of the direct and indirect costs, both short- and long-term. TCO attempts to answer the basic question "What does it cost to own and manage a given technology over its entire lifespan in the organization?" The TCO approach was introduced by the Gartner consulting company in the 1980s.

The total cost of ownership model can be applied in lots of different ways. Many TCO studies have been done, for example, that compare such broad approaches as Linux versus Windows servers or open source versus commercial software.

While TCO can be used by large organizations to help them save millions of dollars when implementing technologies, it can also be used more informally by small organizations as they make decisions about smallerscale projects. Trying to anticipate all the cost implications of a technology-related project is common sense. TCO just reminds us to think broadly and long-term. You don't have to be an economist to incorporate this broader view into the budget projections of a library technology project proposal.

With this broad and long-term cost model in mind, we can lay out the cost components of a typical library technology project. Not all of these considerations will apply to every project, and there might be others not mentioned here that may apply to yours.

Capital expenditures-The most obvious costs of a project involve the upfront capital outlay needed for hardware and software. These expenses can vary tremendously based on whether you need to license commercial software or if there's an open source application available. Will you need to buy new hardware for the project or do you have adequate equipment already on hand? Projects that depend on open source software will have much lower upfront capital costs but may have significantly higher personnel costs as the library takes on full responsibility for the installation, configuration, and maintenance of the software. Given the tradeoffs, it's important to be sure that open source software is implemented on its functional and technical merits and not on the assumption that it will automatically result in lower expenditures.

Recurring costs-Acquiring software rarely involves a single, one-time payment. Most commercial software also requires an annual maintenance fee to make it eligible for product support, bug fixes, and ongoing enhancements. Most library automation vendors charge maintenance fees that are around 15 percent of the amount of the original software license. The annual maintenance fee may also be subject to inflationary increases. Even at a steady 15 percent, the cost of maintenance exceeds the original license price after 7 years.

Direct personnel expenses-All technology projects have significant personnel costs. While a library may not end up hiring new people to carry out a new project, the efforts of the current staff will be diverted from other activities. Personnel costs involve those who implement the systems and those who operate them.

When assessing the cost of a technology project, one of the key considerations is the amount of time required for technical staff. Since libraries compete with the business sector for highly skilled technical personnel, these positions tend to command higher salaries.

Software development-Projects that include the original development of software will naturally have the highest expenditures for technical personnel. A full-blown development cycle that consists of high-level design, articulating functional requirements, writing program code, testing, revising, receding, and retesting can be long, complex, and expensive. A library would only take on this level of effort when no existing software is available to meet the need and if the project returns a very high value.

One of the common oversights is planning only for the actual programming time. The abstract design and quality-control portions of a software development project generally involve as much, if not more, time and cost as writing the program code. You cannot think of these projects as one-time efforts. In the same way we expect commercial companies to deliver bug fixes, enhancements, and upgrades over time, so should local developers properly maintain their software.

Systems administration-Whether the software was written locally or purchased commercially, it will require some care and feeding by technical staff. Initially, the software and all of its prerequisites will need to be installed, configured, and most likely customized. If the software resides on its own hardware, someone will need to install the operating system and apply necessary patches and updates. Any server operating system, whether it's in the Microsoft Windows server family, some flavor of UNIX, or Linux, will need ongoing attention from a systems administrator. Operating systems go through major upgrades every few years, and minor updates and patches must be installed continually.

If existing staff members can absorb the systems administration tasks, then this cost can be fairly minimal. But the effect is cumulative. Technical staff cannot absorb new projects indefinitely. Incremental increases in workload will eventually require an increase in technical staff.

Prerequisite components-Many applications will need supporting components such as a back-end database like MySQL or Oracle and may require you to install a Web server such as Apache. Simpler applications may need a scripting language like Perl or PHP. Applications written in Java will require the Java run-time environment or a full Java development kit. They may also need a Java application server such as Apache Tomcat or a SOAP layer for Web services, such as Axis. Regardless of whether these components are available as open source free software or as commercial products, a competent systems administrator will need to attend to their installation and maintenance.

Database administration-One of the aspects most often overlooked in project planning is database administration. For applications that deal with large amounts of data (and what library applications don't?), it's important to ensure that the database environment is well-managed and has been configured and optimized according to the application's requirements. Depending on the project's scale and complexity, the costs in this area range from negligible to substantial.

Many small projects use database systems such as Microsoft Access or the open source MySQL database. These are among the databases that can be implemented and managed by most systems administrators without undergoing special training.

Higher-end databases such as Oracle, DB2, or even Microsoft SQL Server tend to require deeper expertise and are typically managed by a database administrator (DBA) who has specialized training in that particular database environment. Experienced DBAs find themselves in high demand in the business sector and accordingly can have high salary requirements. Organizations that have a number of enterprisewide applications on a particular database platform tend to employ one or more full-time DBAs. Those with more modest needs may expect other technical staff to be trained as database administrators and to incorporate these database responsibilities into their workloads.

Often during the implementation of a project, librarians will bring in a professional DBA as a consultant rather than train or hire a permanent staff member dedicated to this area of specialization. Only libraries with very large systems departments tend to employ their own DBAs. Instead, many libraries rely on their higher-level IT organization for assistance in this area or follow a more informal approach to database administration.

Content and metadata-Implementing the technical components often just scratches the surface of a project's scope. It's important to give due consideration to the efforts needed to create and maintain the content as well as to the metadata that will be involved. Implementing the hardware and software alone is like creating an empty vessel. The real benefits, and often the lion's share of the work, come from filling the vessel with content and describing that content. Although some projects that have a finite collection may require a one-time content-creation phase, it's far more typical for this work to go on indefinitely. These projects need a permanent reallocation of staff resources.

Training-The execution of a project will likely require many different training components. Technical staff may need to be trained on installing and configuring the system or on managing its underlying components. It's important to be sure that special training is provided when needed and to give the technical staff the necessary time to learn the skills involved. For projects where a broader group of library staff members will be using new software and systems, it's usually necessary to develop a training program to ensure that everyone involved has the opportunity to learn the procedures defined by the project and to gain knowledge on how to operate the software.

Indirect costs-Anyone who has helped create a grant proposal to submit to a federal agency will be familiar with the concept of indirect costs. This relates to the incremental increases in overhead expenses that are associated with implementing a new project. Such costs include the additional work that's required to administer the project's budget, personnel, and facilities. While it's usually not necessary to calculate indirect costs for most non-grant-funded library technology projects, it's good to be reminded that taking on a new project does mean there will be an incremental increase in the administrative staff's workload.

Plan for obsolescence-All technology components have a limited useful lifetime. The hardware you buy today will be unsatisfactory 5 years from now. Operating systems will go through major transitions as will practically any software component involved. If a project's technical components are expected to remain in place for many years, it's important to plan for the costs of an appropriate replacement cycle.

Set Realistic Expectations

While it's necessary to develop a comprehensive assessment of a technology project's total cost, you shouldn't be easily discouraged either. Setting realistic expectations about the cost and impact of a project strengthens the proposal in the eyes of decision makers and reduces the chances that the project will cause frustration due to unanticipated burdens on library staff.

I think that a project's success depends on having a realistic understanding of its full impact on the organization. How much money will it cost? And how much time of how many people will be involved? It's easy to enter into a project with a superficial or incomplete assessment of the resources required. This can cause strain in the organization due to unanticipated financial costs and an unexpected impact on personnel. If an initiative's proponents have a broad view of all the costs involved and produce a project proposal accordingly, it's likely to be better received by the decision makers than a proposal that suggests a lower cost but upon closer inspection leaves out important considerations. Library administrators generally appreciate a candid and complete proposal rather than one that attempts to downplay the investments required.

Whether you're considering a largescale and well-funded initiative or you have to carry out a small project on a shoestring budget, taking into consideration all of the human and financial cost factors and writing an all-inclusive plan will increase its chances of success.

Publication Year:2006
Type of Material:Article
LanguageEnglish
Published in: Computers in Libraries
Publication Info:Volume 26 Number 6
Issue:June 2006
Page(s):26-29
Publisher:Information Today
Place of Publication:Medford, NJ
Notes:Systems Librarian Column
ISBN:1041-7915
Permalink: http://www.librarytechnology.org/ltg-displaytext.pl?RC=12054
Record Number:12054
Last Update:2012-12-29 14:06:47
Date Created:0000-00-00 00:00:00