What Determines IT Spending Priorities?

Global spending on Information Technology (IT) continues to grow and is expected to reach $1.66 trillion in 2009.² In addition, IT represents a large (40–45%) and stable share (in nominal dollars) of firm spending on equipment and software.⁴ As CIOs and IT managers attempt to budget these large IT expenditures they must help the business align IT spending with business strategy and prioritize IT investments in conjunction with business goals.⁵ When prioritizing IT spending managers often compare their own IT spending priorities with those of other firms. That is,

“Every year, CIOs and their finance people get prepared for the following question from their CEOs: “How does our IT spending compare with our competitors?”³

In an effort to help firms benchmark their IT spending priorities, we surveyed 1,495 business leaders in Alabama, Arizona, Colorado, and Texas during the 4^th Quarter of 2005 to examine the firms’ IT spending priorities across business functions (such as, administration; production and distribution; customer relationship management; research and design; managerial decision-making; and security) and the impact of firm and industry characteristics on these priorities. While previous studies examine IT spending from different perspectives,^a we are not aware of any work that examines the factors that determine how firms allocate IT expenditures across business functions.

Empirical analysis of this survey data shows that the highest IT spending priorities of the respondents are in the areas of administration and production and distribution while the lowest priorities are in the areas of research and development (R&D) and security. In addition, the analysis shows that factors such as industry type, firm size, and perceptions of the impact of past IT investments on product quality and revenue may differentially affect the allocation of IT expenditures across business function categories. For example, the results show that:

• Service firms are more likely to rank IT spending in support of R&D and security as their highest priority.
• Small firms are more likely to rank IT spending in support of security as their lowest priority.
• Firms that have leveraged past IT expenditures to increase product/service quality are more likely to rank customer relationship management (CRM) as their highest IT priority.
• Firms that have leveraged past IT expenditures to increase revenues are more likely to rank managerial decision-making as their highest IT priority.

Back to Top

The Survey Instrument and Description of Respondents

The Compass on Business initiative is a collaboration between Compass Bank and four universities: University of Alabama (Center for Business and Economic Research), University of Arizona (Eller College of Management), University of Colorado (Leeds School of Business, Business Research Division), and University of Texas at Austin (IC2 Institute). Compass on Business provides business leaders with insight into economic news and business trends on local, regional, and national levels. One component of this initiative is the quarterly Business Leaders Confidence Index® (BLCI) panel survey. The BLCI is an online survey (accessed via www.blci.com) that asks business leaders in four states (Alabama, Arizona, Colorado, and Texas) six questions about next-quarter expectations on sales, profits, hiring, expenditures, and state/national economic outlook. The survey also asks respondents for firm information such as industry type, annual sales, and number of employees. In addition, each quarter the BLCI survey contains a set of additional questions on a topical issue. Examples include questions about the impact of employee compensation, energy prices, and hurricane Katrina on firm performance.

In Fall 2005, Compass on Business invited the authors to develop a short set of topical questions examining firm IT investments. In response, we developed a set of topical questions (see Figure 1) that asked respondents about their future IT spending priorities and about their perceptions of the economic impact of past IT spending on intermediate performance measures such as quality, costs, and revenues. These questions, which were based on theoretical relationships derived in our previous research on IT value,^9,11,12 were included in the BLCI survey for the 4^th Quarter of 2005.

Of the panelists registered for the BLCI panel, 1,637 responded to the survey in the 4th Quarter of 2005. After eliminating observations with missing data 1,495 observations were used in our analysis. The distribution of these observations across industry categories, based on the North American Industry Classification System (NAICS) is presented in Table 1. For purposes of analysis we further classify the first 8 NAICS categories (1-8) as non-service industries and the last 12 NAICS categories (9-20) as service industries.

We also classify firms as either small (fewer than 100 employees), mid-size (between 100 to 1,000 employees), or large (1,000 or more employees). Sales data was excluded from our analysis due to its high correlation with number of employees (ρ = 0.72, p-value =.000). Table 2 presents a cross tabulation of the respondents by firm size and industry type.

Back to Top

Data Description

In topical question 1 each respondent ranked the firm’s planned 2006 IT spending in six business function categories. In this study we derived the business function categories based on our previous theoretical work on IT value, discussions with IT academics, and feedback from IT managers (based on pilot surveys and focus groups). Although there may be many different ways to categorize business functions we believe this categorization is reasonable and provides many critical insights for IT managers, as we will see. Figure 2 presents a graphical view of the rankings distribution and a tabular view of the overall rank of IT expenditures (based on each category’s average rank) and the percentage of respondents that ranked each business function first and last.

Based on a series of paired t-tests, we confirm that the average rank of each category is significantly different from the category ranked below it. Administration ranks as the highest planned IT expenditure (see Figure 2). New regulatory requirements on firm operations may be one important reason that firms rank administration as their highest IT expenditure. For example, the Sarbanes-Oxley Act requires all public companies to substantiate financial statements with proof of the procedures and controls in place. As another example, Basel II, an accord among bank supervisors and central bankers from 13 countries, specifies international standards for measuring the adequacy of a bank’s capital. Firms must make significant IT investments in administration in order to comply with these and other regulations. In fact, firms will spend $27.3 billion in total compliance spending in 2007, with $6 billion allocated to compliance with the Sarbanes-Oxley Act alone.⁸

Security ranks as the lowest planned IT expenditure (see Figure 2). This result may concern consumers and regulators given the corporate security failures that occurred just prior to the survey. For example, in late 2004 Bank of America lost unencrypted computer backup tapes containing personal information (including credit card information and social security numbers) belonging to 1.2 million federal employees.¹ In October 2004 Choice-Point, a commercial data broker, inadvertently sold personal and financial records of 145,000 consumers to fraud artists.¹ In addition, in June 2005 Visa USA and American Express terminated their contract with CardSystems Solutions when hackers accessed 40 million credit card numbers through their system. Despite these and other security failures, Figure 2 suggests that firms plan to spend the least (relative to other business functions) on protecting company investments, assets, and customer information. While some firms may argue that spending on security is limited due to budget constraints and uncertain market conditions, one might expect that security would have a higher priority given the disastrous consequences associated with security failures and the potential strategic advantage of being viewed as a highly secure firm.

In addition to prioritizing IT expenditures, respondents were asked to what extent their company’s total IT expenditures over the past 3 years changed the cost and quality of their products/services and the revenues of the firm (see Figure 3).

• 39% reported that IT expenditures have not changed their product/service costs. However, almost equal numbers reported that IT expenditures have either increased (32%) or decreased (29%) their costs.
• 71% reported that IT expenditures led to improvements in product and service quality while 25% reported no change in quality.
• 51% reported that IT expenditures resulted in higher revenues while 45% reported no change in revenues.

These results taken together may suggest that firms are leveraging many of their IT expenditures not simply to reduce product costs, but also to improve revenues through product quality enhancements, sometimes at the expense of higher total costs as argued in Thatcher and Pingry.^9,10,11,12

Back to Top

Empirical Model

Having discussed the descriptive statistics, we now examine the impact of firm characteristics and perceptions of past IT performance on the ranking of IT expenditure priorities. Multinomial logistic regression is used to model the probability of discrete outcomes [such as, selecting the best software to acquire].⁶ This method is used to measure the extent to which change in the values of the independent variables may increase or decrease the predicted probability of the event outcome. Here we construct two multinomial models. The first model uses the independent variables, firm characteristics and the perceptions of past IT performance, to predict the business function in which the firm plans to have the highest IT expenditure in 2006. The second model uses the same independent variables to predict the business function in which the firm plans to have the lowest IT expenditure in 2006. These models allow us to measure the impact of a change in each independent variable on the probability a firm ranks a particular business function first and on the probability a firm ranks it last.

In our survey results, there are six business functions (such as, j = 0, 1,…, 5), one of which each respondent may select as the highest (or lowest) IT expenditure in 2006. The model for IT expenditure choice is

for business function j and respondent i. In multinomial logistic regression one value of the dependent variable is designated as the reference category (j = 0). Then, the model of IT expenditure choice becomes,

Differentiating Equation (2) leads to the marginal effects of the independent variables on the probability of selecting business function j as the highest (or lowest) IT expenditure,

The marginal effects are presented in Table 3 and are the basis of the analysis that follows.

Back to Top

Data Analysis

Earlier we noted that the majority of survey respondents believe that their firms’ portfolio of IT investments over the past three years has led to improvements in product quality and firm revenues. This suggests that firms may be aligning their IT investments with business strategy to improve consumer value and firm performance. Interestingly, firms that have leveraged past IT investments to improve product quality are more likely to rank CRM and less likely to rank administration as their highest planned IT expenditure. CRM software manages how a firm interacts with its customers. As such, firms may use CRM software to analyze customer data, collected from call centers or sales force interactions, to improve customer service quality.

In addition, firms that have leveraged past IT investments to improve firm revenues are more likely to rank managerial decision-making and less likely to rank security as their highest planned IT expenditure. IT investments in support of security are typically not considered strategic and do little to improve firm revenues. However, decision support systems, data mining tools, and other business analytic software help firms analyze business data to forecast market trends, identify patterns, and make better business decisions. Firms may use these business analytic tools to more effectively segment customers and engage in targeted marketing strategies, leading to more revenues.

Back to Top

Firm size

Table 4 suggests that small- and mid-size firms rank administration as a higher priority than large firms, while they rank R&D and production and distribution, which are important strategic business functions, as relatively lower priorities. In small and mid-size firms, IT departments are still often recognized as cost centers. Only one-third of IT spending in these firms is strategic and aimed at improving business performance.⁷ Instead of allocating their limited IT resources to strategic business functions, these firms spend most of their IT budget on the fundamental administration (such as, payroll, email, compliance, and other back-office solutions) or on supporting infrastructure (such as, network and servers). This observation is consistent with our results and suggests that small- and mid-size firms may be under-investing in strategic IT, failing to align IT and business strategy, and, therefore, not realizing the quality and revenue improvements enabled by such alignment.

In addition, small firms tend to rank security as a lower priority than large firms. Some argue that small firms often ignore data security issues. One reason may be that developing effective security measures is very costly for small firms with limited budgets. Another reason is that some small firms wrongly believe that hackers are only interested in large, high profile firms. However, hackers often target small firms as testing grounds or launching pads for larger attacks. Therefore, small firms may be more vulnerable to security beaches than they think, leading to under-investment in security applications (relative to other functions).

Service Firms. Six of the seventeen significant results presented in Table 4 lie in the service column. Here we summarize two important results derived from this analysis. Earlier we noted that, on average, security was ranked as the lowest planned IT expenditure by respondents (see Figure 2). Interestingly, Table 4 shows that service firms are more (less) likely to rank IT investments in security as the highest (lowest) IT expenditure. This result may be due to recent regulation that increasingly requires firms, especially those in the areas of health care, banking, and financial services, to protect customer data. For example, the Gramm-Leach-Bliley Act (GLBA) requires banks and financial services firms to protect customer data while the Health Insurance Portability and Accountability Act (HIPAA) requires health providers to secure electronic patient information. These regulatory requirements may lead to an increase in IT expenditures by service firms in support of security. However, we note that despite these marginal effects service firms still only rank security as the fifth priority compared to non-service firms which rank it last (see Table 5).

In addition, the service industry has been a hotbed of product customization based on research using data gathering and data mining tools. This focus on customization to support changing customer needs may be one reason that service firms are more likely to rank R&D first than are non-service firms (see Tables 3 and 4). In fact, 15.44% of service firms ranked R&D first compared to 9.74% of non-service firms (see Table 5).

State Differences. In three cases, the state in which a firm operates significantly affects the firm’s IT expenditure priorities. Specifically, when compared to Texas, firms in Alabama, Arizona, and Colorado are all more likely to rank administration as their highest IT expenditure. On the other hand, when compared to Texas firms, firms in these same states are less likely to rank production and distribution as their highest IT expenditure. A final interesting observation is that firms in Arizona are less likely to rank security as their highest IT expenditure when compared to the other states. Although it is difficult to determine the source of these state differences based on this survey data, it may reflect state-specific differences in the business environment (such as, regulation).

Back to Top

Conclusion

Based on survey results from 1,495 business leaders we have derived some important insights into how factors such as industry type, firm size, state location, and past IT performance affect a firm’s allocation of IT expenditures across business functions. Specifically, the survey data shows that the highest IT spending priorities of the respondents are in the areas of administration and production and distribution while the lowest priorities are in the areas of R&D and security.

Back to Top

Back to Top

Back to Top

Back to Top

Back to Top

Back to top