Costs and Benefits of GIS in Business
James Pick, University of Redlands, USA
Abstract
This chapter examines the costs and benefits of geographic information systems (GIS).
It focuses on the research questions of what components to include in GIS cost-benefit (C-B) analysis, what distinguishes GIS C-B analysis from non-spatial C-B analysis, what methods to use, and how to invest in GIS systems in order to obtain net payoffs over time. It categorizes costs and benefits of GISs. It considers the topics of systems analysis sub-steps in cost-benefit analysis, the feasibility decision, stakeholders and externalities, and the importance of timing and timeliness in investing in GIS and assessing payoffs.
It examines the C-B aspects of a well-known GIS case study of Sears Roebuck’s delivery system. The literature on the value of investment in IT and productivity paradox is analyzed for its relevance to GIS investment. The major findings are, first, that the costs and benefits of a GIS can be estimated through modifications of standard non-spatial IS methods. Second, the key factors that differentiate GIS cost-benefit analysis from that of non-spatial IS are more extensive analyses of the costs of data acquisition, need to pro-rate the GIS costs and benefits for tightly linked combinations of GIS and other systems and technologies, and need for improved techniques to estimate the costs and benefits of the GIS’s visualization features.
Introduction
Geographical Information Systems (GISs) have become an important tool for government and business decision-making (Huxhold, 1991, 1995; Grimshaw, 2000; Tomlinson, 2003;
Costs and Benefits of GIS in Business 57
Clarke, 2003). One early definition of GIS is the following: “The purpose of a traditional GIS is first and foremost spatial analysis… Capabilities of analyses typically support decision-making for specific projects and/or limited geographic areas” (Exler, 1988). GIS is more than mere mapping, extending much further in its capabilities to analyze spatial information through such techniques as overlays, queries, modeling, statistical compari- sons, and optimization (Mitchell, 1999; Clarke, 2003; Greene & Stager, 2003).
GIS performs spatial analysis based on data and boundary files stored in databases to support decision-making (Huxhold, 1991; Murphy, 1995; Jarupathirun & Zapedi, 2001, 2004). GISs provide mapping and analysis for marketing, transportation, logistics, resource exploration, siting, and other business sectors (Harder, 1997; Grimshaw, 2000;
Boyles, 2002). GIS can take advantage of spatial factors to improve response times, decide more efficiently on locations, optimize movements of goods and services, market more effectively, and gain enhanced knowledge of routing, siting, and territories. GIS can be linked to the Internet to support virtual applications. The core of the GIS market, consisting of software, services, data, and hardware, is estimated to total $1.75 billion in 2003 (Directions Magazine, 2003). Projections indicate this GIS market grew by 8 percent during 2003, which occurs as the U.S. came out of recession (Directions Magazine, 2003).
Cost benefit analysis for GIS is done for three main reasons (King & Schrems, 1978).
1. Cost-benefit analysis assists in planning for an organization. Planning involves tradeoffs between competing demand for organizational investment and resources.
C-B analysis can help in deciding between competing demands.
2. Cost-benefit analysis is useful in auditing. The organization may decide after a GIS system has been put into effect to perform a C-B analysis retrospectively, to assess what occurred.
3. To prepare and support participants in political decision-making (King & Schrems, 1978). This is less formal and more rapid than the planning uses in numbers 1 and 2. It is mostly done with less complete information, but is more commonplace than the other reasons (King & Schrems, 1978).
The objective of this chapter is to answer three research questions: (1) How can the costs and benefits of a GIS be estimated? (2) What differentiates cost-benefit analysis of GIS from standard cost-benefit analysis of information systems? and (3) What are the appropriate data and instruments to measure costs and benefits of GIS? In answering the questions, the chapter considers financial, technical, institutional, and integration costs and benefits. It also looks at the value of IT investment methods, to assess their relevance for GIS.
Several prior research studies have examined cost-benefit analysis of GIS and IS.
Obermeyer (1999) summarized several methods to analyze costs and benefits of GIS- based systems. Tomlinson (2003) discussed the principles and rationale for cost-benefit analysis of GIS, emphasizing certain features that are distinctive about spatial analysis.
Taking a wide view of costs and benefits, on the cost side he added to the usual items the broader categories of liability, software interfaces, and communications. On the benefit side, he added better timing of projects, richer information to users, improved
58 Pick
organizational workflow, and more effective expenditures by the firm. However, often firms must wait over the lifetime of a project to receive the benefits (Tomlinson, 2003).
It may take a GIS five or more years to reach the breakeven point where cumulative benefits start to exceed cumulative costs (Tomlinson, 2003). Huxhold & Levinsohn (1995) discussed the benefits of GIS from a project management perspective. There is also research that focuses on the investment in, and value derived from, IT for different units of analysis, such as the project, firm, and industry (Ahituv & Neumann, 1990).
A related line of literature has involved the information technology productivity paradox, which refers to studies of the investments in, and payoffs from information technology (IT). Some of this research has indicated low or no payoff (Brynjolfsson, 1993;
Strassmann, 1997, 1999; Lucas, 1999). Several conceptual frameworks exist for the productivity paradox, including normative value, realistic value, and perceived value (Ahituv, 1980, 1989). Recent literature in this area has been more likely to conclude that there are net benefits from IT investment (Ragowsky et al., 2000; Deveraj & Kohli, 2002).
One reason for a more optimistic picture is that methods and data collection have improved, so that benefits that might have been missed before can now be accounted for.
In addition, studies are better able to judge the appropriate timeframes over which to measure the benefits, taking into account lagged effects, and to disaggregate the unit of analysis into smaller units (Brynjolfsson, 1993; Ragowsky et al., 2000; Deveraj & Kohli, 2002; Navarrete & Pick, 2002).
The specific topics covered in this chapter are: (1) Summary of costs and benefits, (2) Systems analysis sub-steps for cost-benefit analysis, (3) Comparison of costs and benefits, (4) Analysis steps in cost-benefit analysis, (5) Stakeholders and externalities, (6) The case of GIS costs and benefits at Sears Roebuck Delivery, (7) Value of IT investments: its relevance to GIS, and (8) Conclusion.
The methodology utilized in the present chapter is critical evaluation of whether or not non-spatial IS cost-benefit techniques can be applied to GIS, and the analysis of the Sears case example. The chapter is broad and draws on a variety of literature. It does not perform empirical analysis with real-world data to determine net positive benefits.
However, the chapter may be useful as a framework to researchers conducting such empirical studies.
Setting the Context for Cost-Benefit Analysis
A GIS cost-benefit analysis must consider and clarify at the beginning its broad context, especially four factors (King & Schrems, 1978):
1. Statement of purpose. This statement indicates whether the C-B analysis is being utilized directly to make a decision, to provide background data for a decision, or to politically influence decision-making (King & Schrems, 1978). It is important to state the purpose, since the methods, quality of the data, and reporting of findings, among other things, differ by purpose.
2. Time simultaneity. The C-B analysis must indicate whether the C-B analysis is prospective to a future GIS system, for a current GIS system, or retrospective. All three are useful in certain circumstances.
Costs and Benefits of GIS in Business 59
3. Scope. The C-B analysis may be comprehensive in examining all possible costs and benefits. On the other hand, it may be severely limited in scope: for instance, it may only manipulate a single cost item, in a sensitivity analysis, to see if change in that item affects the benefit outcomes.
4. Criterion. The last contextual factor is the method that is used to compare the costs and benefits, after they have been compiled. This may be a quantitative index measure, graphical comparison, or involve a lengthier model. The criterion needs also to state whether or not the values of the costs and benefits are to take into account inflation and opportunity costs through present-value calculations.
Among the challenges in GIS cost-benefit analysis is that GIS usually has higher costs than for conventional information systems (ISs), due to its considerable data acquisition and data management. This is because a GIS is based on both attribute data and spatial data (Huxhold, 1991; Clarke, 2001). The extra time and effort, versus non-spatial IT, stems from the need to do the following:
• Gather boundary data and associated attributes
• Convert the data to digital form
• Design or configure topological data structures and link non-spatial with spatial data
• Maintain GIS boundary files and data
Estimates indicate data collection for GIS may constitute 65 to 80 percent of the total cost for conventional systems development/implementation (Huxhold & Levinson, 1995;
Obermeyer, 1999; Tomlinson, 2003). Further, the attribute and digital boundary data need to be linked together. These linkage tasks add to costs, relative to non-spatial IS.
Another difference has to do with GIS’s feature of visualization (Jarupathirun & Zahedi, 2004). In this respect, GIS is comparable to a multimedia business application, which may have higher costs, due to its visual aspects. Because the benefits of visualization are predominantly intangible, GIS tends to have a higher proportion of intangible costs and benefits than a non-spatial IS. Another distinctive aspect of GIS is that it tends to be more linked or coupled with other software systems and technologies than is normally present in IS applications. Among the systems and technologies with which GIS is often interfaced are global positioning systems (GPS), remote sensing (Meeks & Dasgupta, 2003, 2004), and marketing information systems (Allaway, Murphy, & Berkowitz, 2005).
As seen in Figure 1, because of this linking together of several types of systems and technologies, the cost and benefit calculation for any one of them may be more difficult.
Figure 1 demonstrates that the assessment of costs and benefits of a simple, uncoupled GIS system can be done by performing an assessment of that system’s costs and benefits, followed by a cost-benefit comparison such as break-even analysis. However, the cost- benefit analysis becomes more complicated for a GIS that is closely linked with another system or technology, such as point-of-sale, remote sensing, or GPS. For linked systems, it is difficult to disaggregate the costs and benefits of the GIS from those of the other parts.
60 Pick
Summary of Costs and Benefits of a Geographic Information System
This section discusses the categories of costs and benefits of a geographic information system. Costs and benefits may be divided into tangible — i.e., able to be converted into monetary amounts — and intangible — i.e., not convertible to monetary values (King &
Schrems, 1978). Costs for information systems including GIS are predominantly tangible, while benefits are a mixture of tangible and intangible. For instance, the cost of a GIS managerial employee can be estimated by the tangible value of his/her salary and job benefits. However, the most significant benefit of the employee is his/her effective leadership and decision-making, outcomes difficult if not impossible to convert to dollar amounts.
Figure 1. Comparison of Cost-Benefit for a Simple GIS System to a GIS System Closely Coupled with Another System or Technology
Cost-benefit analysis for a simple uncoupled GIS
Cost-benefit analysis for a GIS system closely coupled with another system or technology
Costs and Benefits of GIS in Business 61
Costs
The costs of a GIS can be classified into the categories given in Table 1. These costs are all tangible and are possible to estimate. However, some problems occur in accurately accounting for the costs, and by the risk of exclusion of some of the full range of costs.
It is important to avoid the many common errors prevalent in cost accounting (King &
Schrems, 1978), such as not identifying hidden costs, counting costs twice, or omitting important costs. An example of hidden costs would be costs located at other places in the organization that are not being counted for the GIS. For instance, the development of a GIS system depends on the ideas of several top managers who spend significant time with the applications development team. However, the overhead of the time spent by the top managers with the team is not included in the costs. Omitted costs are ones that are not obvious, but are in fact dedicated to the project. For example, space, site, and utility costs are commonly omitted, but may be important, particularly in expensive locations, such as midtown Manhattan.
Benefits
Benefits of GIS and of ISs in general are more difficult to measure than costs (King &
Schrems, 1978; Obermeyer, 2000; Tomlinson, 2003). The reason is that benefits often accrue in the form of a more informed, ready, efficient, and high-performance organiza- tion, which is hard to measure, since there are many beneficiaries, time lags, and intervening causes. A deeper problem is the benefit of information value may increase, but the value of information is difficult to measure (Ahituv, 1989). The difficulty stems
Table 1. Tangible Costs of a GIS
ã Hardware
ã Software
ã Data collection
ã Transformation of manual maps and data into digital format
ã Maintenance costs for hardware and software
ã Maintenance of data
ã Supplies
ã Design and construction of databases
ã Hiring more staff
ã Training present staff
ã Outsourcing (e.g., GIS applications programming)
ã Consulting
ã Licensing
ã Communications interfaces and networks
ã Space, site, and utilities
Modified and expanded from Huxhold (1991), Obermeyer (1999), Tomlinson (2003)
62 Pick
from information value being dependent on its timeliness, users, and what decisions it is influencing. Another problem with measuring benefits has to do with quantity versus quality. If a GIS system leads to a higher quantity of maps being produced than the former manual system, the key related question is how does the quality of the GIS maps compare to that of the manual ones. Hopefully, the GIS-based maps would have higher quality, but that may not be so, in which case the output gains in quantity of maps would be offset by reduced quality.
In spite of the problems, benefits can be, and are calculated and utilized. Practical suggestions for evaluating benefits include the following.
1. Larger benefits may be disaggregated into smaller pieces, which may be more amenable to quantification. Consider an example. A truck transport company has identified the benefit of better control of its fleet by its GIS. However, the overall control cannot be made tangible. Nevertheless, if control is broken down into a set of small control items, such as local supervisory knowledge of trucks arriving in each city, supervisory knowledge of trucks loading in each city, etc., control for the small items is now more amenable to assignment of dollar values, although it is still not easy.
2. The C-B analysis can be restricted to only tangible costs and tangible benefits (King & Schrem, 1978). If that result indicates net benefit, then the intangible benefits may be regarded as an added plus. Another variation on this approach is to perform a break-even analysis for a future time point, just restricted to tangibles (King & Schrem, 1978). If costs exceed benefits, the difference of benefits minus costs can be compared against intangible benefits. They may be close enough to yield a compelling argument of justification.
Unit of Analysis
The difficulty in measuring the costs and benefits of GIS varies by the unit of analysis (Obermeyer, 1999; Tomlinson, 2003). Among the units of analysis for GIS are the following:
• Industry
• Company
• Department or division
• Project
• Individual
Most commonly, the unit of analysis is the company or project. For units with small scope, such as a single-user desktop project, costs and benefits may be more readily estimated. For them, there is limited integration with other systems and technologies;
intervening factors are reduced, and the external environment is not as influential. On the other hand, if the unit of analysis is a corporate-wide GIS system, it may be challenging
Costs and Benefits of GIS in Business 63
to separate its costs and benefits from those of other systems inside the company, from inter-organizational systems, or from the outside environment. Furthermore, for enter- prise-wide GIS, the attribute data are commonly shared with other company systems, such as marketing, or with the firm’s enterprise resource planning (ERP) system. The shared aspect of the business data complicates the separation of GIS’s costs and benefits.
At the firm level, the task of separating the benefits of GIS from other intervening factors is even more prevalent, since extraneous influences become more important, such as economic impacts, competitors, and government policy changes. An illustration of the challenges of correlating returns on IT investment at the firm level was seen in a longitudinal study at the firm level for the Mexican Banking industry from 1981 to 1992 (Navarrete & Pick, 2002). The economic cycle during this period was seen to be influential Table 2. Benefits of GIS (Tangible and Intangible)
T a n g i b l e
ã Salary and benefits lowering from reducing the workforce
ã Cost reduction (through employees performing their tasks more efficiently)
ã Cost avoidance in the future (projected greater workload per employee)
ã Expansion of revenues (achieved through improved data quality, improved efficiency)
ã Improved productivity
ã Improved performance
ã Higher value of assets
I n t a n g i b l e
ã Improved decision-making (at issue, how is the tangible value of better decisions estimated.
The data for decision-making may be faulty. GIS can only contribute so much if the data are faulty.)
ã Effectiveness of managers and executives
ã Reaching strategic objectives
ã Environmental scanning
ã Speed and timeliness of information
ã Volume and quality of information
ã Better capability to sell products (CDs, web services, manual maps)
ã Improved collections of money
ã Identification of missing revenue sources (e.g., in government, identification of properties not being taxed)
ã Better operational efficiency and workflow
ã Better utilization of assets
ã Reduced error
ã Reduced liability (e.g., GIS for security monitoring)
ã External benefits (i.e., benefits to organizations other than the one implementing the GIS; an example would be benefits from a marketing GIS product to the customer buying the product)
Modified and expanded from Huxhold (1991), Obermeyer (1999), Stein & Nasib (1997), Deveraj & Kohli (2002), Tomlinson (2003)
64 Pick
on the level of IT investment for a given year. Availability of funds for IT investment depended not only on bank cash flow, but also on the outside factors of economic and market conditions of interest rates, disposable incomes of customers, banking-industry competitive pressures, and pricing of hardware and software products.
Since GIS technology is fast moving, there may not be precedents for modeling or benchmarking some future costs or benefits. For instance, hand-held GIS products such as ArcPad have appeared in the last five years, and consist of a limited GIS software package, a keypad or other means to enter data, a small display screen, and cabling connectors for information import and export (ESRI, 2003). However, given that these hand-held devices are early in their product cycle, there may not be benchmarking studies available on their performance, reliability, and ease of use (Day & Schoemaker, 2000). As new products, their costs may be unstable. On a long-term basis, technological changes for them may be hard to evaluate, since some new technology products may fail, while others remain relatively stable, and yet others are redesigned rapidly (Day & Schoemaker, 2000; Doering & Parayre, 2000).
Systems Analysis: Sub-Steps in Cost-Benefit Analysis for GIS
In systems analysis and design, the analysis stage includes a feasibility study that contains cost-benefit analysis as a part of it (Ahituv & Neumann, 1990; Satzinger, Jackson, & Burd, 2002).
The following are the standard cost-benefit sub-steps:
1. Develop an overall plan for the C-B analysis.
2. Decide on the analyst or analyst team.
3. Determine the alternative C-B analyses to be conducted.
4. Determine all the material factors for costs and all the material factors for benefits.
5. For each tangible factor, decide how it will be measured.
6. Measure the costs and benefits.
7. Compare the cost-benefit results over the entire time period of the study. Include summary measures such as break-even point, etc., and criteria to evaluate alterna- tives.
8. Perform a comparative analysis of the alternatives from step 3.
9. Decide on what recommendations to make to management, based on these results.
Present the findings to management (modified from Ahituv & Neumann, 1990; King
& Schrems, 1978).
Before commencing C-B analysis, an analyst must be selected to conduct the analysis
— a choice that is often critical for C-B’s success or failure. There are a number of places