Vous êtes sur la page 1sur 43

Executive Overview This article is a study of 98 small manufacturing firms (average sales $5 million; firms' expenditures about 1 % of sales;

; average number of employees 62), and the effectiveness of their computer use. Thefindings were applicable not just to small companies but, in large measure, to small divisions of largercompanies as well. The key findings showed tbat effectiveness was linked to: 1. Chief executive officer knowledge of computers and active involvement in the computerization efforts.This was clearly the strongest factor. 2. On-site computers (as opposed to a remote service bureau). The Senior Editor was surprised by this item. 3. The coordinated implementation of planning and controls. Equally interesting, effectiveness was not linked to: 1. The use of external programmers. Internal staff, even in these small companies, was just as likely to produce good results. 2. Employees' enthusiasm for computer technology. While a minimal level of acceptance is needed, higher levels of acceptance did not improve effectiveness. 3. Length of time of computer use. The passage of time did not, surprisingly, improve effectiveness. 4. Formal computer training of employees. 5. Planning, unless strong computer controls are also in place. In summary, many of the issues and findings in large organizations are also present in their smaller counterparts. 50 MIS Quarterly/March 1988 Small Business Determinants of Success for Computer Usage in

Smaii Business By: William H. DeLone Kogod College of Business Administration The American University 4400 Massachusetts Avenue N.W. Washington, D.C. 20016 Abstract This field study investigates the factors that affect the successful use of computer-based information systems (OBIS) by small business. The 93 manufacturing firms surveyed had fewer than 300 employees, less than $30 million in annual sales revenues, and had been using computers for at least three months. The principal findings showed that chief executive knowledge of computers and involvement in computerization leads to more successful computer use in small manufacturing firms. The use of on-site computers also has a positive effect on computer success. Keywords: Small business, IS success, computer usage ACM Categories: H.4, J.I, K.4.3, K.6 Introduction With the widespread availability of microcomputers, the cost of small business computer systems has been reduced to a point where nearly all businesses, no matter how small, can afford computer power for their information processing needs. The question for the small firm manager

who is not knowledgeable about computers is: "How might my firm be able to use computers successfully?" This study attempts to identify the factors associated with computer success in small firms. Of the 11 million nonfarm businesses operating in the United States in 1980, 10.8 million firms (98.2 percent) were considered "small" by the Small Business Administration (SBA) size standards. Small businesses accounted for 39 percent of this country's GNP in 1976 and for 48 percent of nongovernment, nonfarm employment (U.S. SBA, 1981,1983). The importance of the small business sector to the United States economy is clear. Computerized systems can help solve small business problems and thereby improve prospects for success. Although business failure rates are not reported by firm size, small businesses are considered to be more risky (Brigham and Smith, 1967; Walker, 1975) and subject to higher failure rates than large businesses (Cochran, 1981; Klatt, 1973). Problems with receivables collections and with inventory control were cited as contributing to 17 percent of the manufacturing business failures studied by Dun and Bradstreet (1981). Generally, small businesses have difficulty maintaining adequate records (Markland, 1972, 1974; Rotch, 1967). These problem areas can be addressed by computerization. Small business computers can also

help to improve service and increase sales, which are important contributions considering that inadequate sales is involved in 59 percent of business failures (Dun and Bradstreet, 1981). Therefore, small business computers have the potential to make important contributions for thousands of small firms. Despite this potential, small businesses should approach computerization cautiously. Lack of computer knowledge on the part of the owner/ manager and lack of computer experience have resulted in all too many misadventures in electronic data processing (Apcar, 1980; Schollhammer and Kuriloff, 1979). Small businesses are financially ill-equipped to absorb such expensive MIS Quarterly/March 1988 51 Small Business mistakes (Charlesworth, 1972; Cohn and Lindberg, 1972; Schollhammer and Kuriloff, 1979). So computers, if managed properly, can contribute to the success of small business operations, but the risks are significant. Guidelines for the successful application of computers are needed. This study investigates the factors that are associated with the successful use of computers in 93 small manufacturing firms in Los Angeles. Research Variables and Hypotheses MIS success As an initial step, this study attempts to develop a meaningful measure of a successful computer-

based information system (CBIS) that is appropriate for small businesses. A review of the MIS literature suggests several measures of success for large organizations. Among the suggested measures are management use of the information system and the impact of the information system on organizational performance. A frequently suggested measure of success is the extent to which the information system (or MS/OR system) is used by management (Cerullo, 1980; Ginzberg, 1981; King and Rodriquez, 1978; Lucas, 1975, 1978; Zmud, 1979). Ein-Dor and Segev (1978,1981) claim that different measures of computer success are mutually dependent and they choose system use as the primary criterion variable for their proposed research framework. If use is the success measure of choice, the issue then becomes "use how and by whom?" (Huysmans, 1970). Actual use as a measure of MIS success only makes sense for voluntary users, as opposed to required users (Lucas, 1978; Weike and Konsynski, 1980). Another way to measure success is to determine the impact of an MIS on individual or organizational performance (Cerullo, 1980; Ein-Dor and Segev, 1978; Hamilton and Chervany, 1981; Kriebel, 1979; Lucas, 1975). Attempts have been made to determine the effect of computer systems on a firm's profits and to compute the firm's return on computer investment (Garrity, 1963;

McKinsey, 1978). A more appropriate success measure for a small business might be the effect of information systems on cash flow (Seibt, 1979; Williams, 1978). The impact of information systems characteristics on profit and/or performance can be isolated in tightly controlled experiments but is difficult to determine in field studies. As Hanes and Ramage (1983, p.84) state: "The problem is that too many other uncontrollable and unmeasurable factors influence organizational performance." The researcher thus needs to determine the extent to which a firm's MIS contributes to the firm's critical business areas or to the firm's "life stream" systems (Couger and Wergin, 1974; Ein-Dor and Segev, 1978; Greenwood, 1981; Rockart, 1979; Senn and Gibson, 1981). MIS benefits to the firm could be represented by the range of meaningful computer applications (McKinsey, 1968; VanLommel and DeBrabander, 1975). Based on previous research conducted in larger organizations and based on field responses from small businesses (DeLone, 1981), two measures were selected as appropriate for the small manufacturing firms in this study. These measures are 1) actual use of computer-generated reports by top management, and 2) the impact that the computer applications are having on the business. The composite impact measure is based on importance and success ratings given to the

firm's applications by the chief executive officer. Success factors The proposed factors to influence CBIS success are chosen on the basis of their perceived importance in the context of a small business setting. The success factors (independent variables) selected for this study are discussed below. Small businesses typically lack specialized knowledge and technical expertise (Charlesworth, 1972; Deeks, 1976; Klatt, 1973; Schollhammer and Kuriloff, 1979). This shortcoming applies equally to knowledge about computers. Lack of understanding about computers is a frequently cited reason for failure of small business computer endeavors (Raysman, 1981) and for failure to consider computer opportunities (Neidleman, 1979; Weber and Tiemeyer, 1981). However, there are ways to overcome these shortcomings. Computer experience, top management involvement (Bourke, 1979; Couger and Wergin, 1974; Greenwood, 1981; Newpeck and Hallbauer, 1981), company-supported EDP 52 MIS Quarterly/March 1988 Small Business training (Bevis, 1979; Heise, 1980; Vazsonyi, 1981; Weber and Tiemeyer, 1981), and the use of external computer expertise (Briggs, 1980; Couger and Wergin, 1974; Greenwood, 1981; Senn and Gibson, 1981) can increase computer knowledge or compensate for lack of knowledge

and thereby improve the chances of successful computer use by small firms. Personnel acceptance of computer systems (Heise, 1980; Newpeck and Hallbauer, 1981), formal planning of computer efforts (Hansen, 1980; Senn and Gibson, 1981), and the implementation of computer controls (Couger and Wergin, 1974) are other factors which are said to affect the success or failure of small business computer systems. Finally, the firm's maturity regarding computer use and, more specifically, type of computer use (in-house, service bureau, etc.) are situational factors that are also likely to affect the success of computer operations. Based on these criteria, the success factors (independent variables) chosen for this field study are: 1. the use of external programming support (EXTERNAL), 2. the level of CBIS planning (PLAN), 3. top management knowledge of computers (EDPKNOW), 4. top management involvement in computerization (INVOLVE), 5. personnel acceptance of computers (ACCEPT), 6. the sophistication of computer controls (CONTROL), 7. the age of computer operations (AGE), 8. the level of computer training (TRAINING), 9. the type of computer use (TYPE). Hypotheses The nine success factors discussed above are

formulated into hypotheses. The hypotheses test whether small manufacturing firms realize a higher level of CBIS success when they have: HI: greater use of external programming support H2: higher levels of CBIS planning H3: a chief executive with greater computer knowledge H4: a chief executive who is more deeply involved in the computerization of applications H5: higher levels of computer acceptance by employees H6: more sophisticated computer controls H7: used their computers for a longer period of time H8: higher levels of computer training for their employees H9: on-site computers (vs. use of computer services) Methodology The hypotheses listed were tested against data collected from small manufacturing firms. The sample of small business computer users was randomly selected from the population of all firms listed in the California Manufacturers Register (1981). These firms were located in the city of Los Angeles, employed less than 300 persons, and earned less than $30 million in sales revenues. The sample for this study was limited to the broad industry category of manufacturing

in order to eliminate industry as a source of variation. This study involved two questionnaires as its data collection instruments. A telephone survey was conducted to determine which randomly selected manufacturing firms had been using computers for at least three months. Both questionnaires were sent to the chief executive of each of the selected manufacturing firms who were found to be using computers. The first questionnaire, the "Chief Executive Ouestionnaire," was to be completed by the company's top executive. It included questions about the success of the firm's computer-based MIS and items relating to computer understanding and experience. The second questionnaire, entitled "Computer Use Ouestionnaire," involved questions pertaining to the success factors under investigation. It was completed by the employee who was administratively responsible for the company's computer processing. The two questionnaires were mailed to 191 firms. Ninety-three firms returned both questionnaires for a response rate of 48.7 percent. The questionnaire responses indicate that these firms averaged 62 employees and $5 million in annual sales revenues, and had been using computers for an average of 48 months. (Medians were used as the measure of average due to positive skewness in most data items.) MIS Quarterly/March 1988 53

Small Business An abbreviated version of the Chief Executive Questionnaire was sent to the companies that did not respond to the survey. Forty-one of the nonrespondents returned this shorter questionnaire. These questionnaires were used to determine whether there were any significant differences in the characteristics of the respondent and nonrespondent firms. Difference tests on each of six items (firm size, type of computer system, age of computer operations, executive computer experience, overall success rating for the computer system, and organizational level of respondent) proved to be nonsignificant. Therefore, a reasonable assumption is that the respondent data is representative of the entire population of small manufacturers which are using computers. Additionally, since 72 percent of the Chief Executive Questionnaires were actually completed by the chief executive, and the remaining 28 percent were completed by the next level of top management (such as controllers and vicepresidents), the responses reflect the opinions of the top executives of the surveyed manufacturers. The two CBIS success measures for this study are entitled USE and IMPACT and are presented in Table 1 as they were measured from the responses on the Chief Executive Questionnaire. Table 1 indicates a brief description of relevant questionnaire items and the average value for

each item from the 93 questionnaires that were returned. The median was selected as the measure of average because many of the variables were positively skewed. For analysis purposes. USE and IMPACT measures which were positively correlated (at the .05 level) were later combined into a single success index. The success factors are presented in Table 2 as they were measured from questionnaire responses. Table 2 includes a brief description of the relevant questionnaire items and the average value for each item over the 93 questionnaires received. Since one purpose of this paper is to develop new theory related to small business computer use, a number of variables have not been previously validated; therefore, nominal scale analysis (multivariate cross-classification analysis) is used to test the hypothesized associations. All variables which are not categorical by nature are classified into two or three categories (low. moderate and high) in preparation for analysis. The category boundaries are set on the basis of variable distributions with the goal of balancing the frequencies in each category. As discussed earlier, the SUCCESS variable (S) is a composite measure of computer report USE by the chief executive and the IMPACT of computer applications in the view of top management. The computer knowledge variable (K) is also a composite measure which combines chief executive exposure

to computers with chief executive formal computer training. Results Multivariate cross-classification analysis was performed to test the research hypotheses. Multivariate cross-classification analysis, or the analysis of multidimensional contingency tables, was based on loglinear models models which are linear in the logarithms of the expected cell Table 1. CBIS Success Measures Success Area USE IMPACT' Questionnaire Item Computer Report Usage (time spent) Computer Report Usage (frequency) Number of Application Systems Application Importance Score Application Success Score Median 4 hours/mo. 10 times/mo. 5 3.57^ 3.25= For analysis. IMPACT was computed as the product of the application importance score and the application success score, summed over all applications (^ Importance; x Successi where N = No. of Applications). "' These scores range from one (lowest) to four (highest). 54 MIS Quarterly/March 1988

Small Business Table 2. Success Factors FACTOR EDPKNQW (K) AGE (G) TRAINING (R) INVOLVE (1) ACCEPT (A) PLAN (P) CONTROL (C) EXTERNAL (E) TYPE (T) Questionnaire Item Years of Computer Experience Executive Computer Training (for 44 execs receiving training) Age of Computer Operations Overall Computer Training for Employees (for 28 firms with training) Chief Executive Interaction with D.P. Manager Number of Computer-related Complaints Planning Score (scale from 0 to 10) Control Score (scale from 0 to 18) Percentage of Applications Developed Externally (55 of 93 firms totally dependent on external software) On-site vs. Service Bureau Median 4 yrs.

30 hrs. 48 mos. 3 hrs/mo. 2 hrs/mo. 1 per mo. 7 12 100% 70% on-site values (Bishop, et al., 1975). The statistical calculations were performed by the BMDP4F computer program (UCLA, 1979). Since only 93 complete responses were received, and since most of the variables were classified to include three categories (low. moderate, high), three-way cross-classification analysis was the highest order analysis that could be performed while avoiding the analysis problems associated with zero and low expected cell frequencies. For each hypothesis, the two-way association of interest was tested in the presence of other independent variables taken one at a time, so that each hypothesis was tested by more than one three-way cross-classification test. The loglinear model was the basis for constructing and performing tests of association between categorical variables. The strategy was to test various simple, restricted models against the general, "saturated" model. This study compared a restricted model to the general loglinear model by comparing goodnessof-

fit statistics for each. The goodness-of-fit statistic that was used for the analysis of the research hypotheses was the likelihood ratio chisquare (Feinberg. 1980) expressed as: G= = 2 = 2 (Observed) log (summation is over all cells) In order to test the significance of an association, the difference between the G= for the 3 variable model without that association and the G^ for the 3 variable model which includes that association was tested over the difference in degrees of freedom for those models. For the associations proposed in this study, the null hypothesis assumed independence (no association) and the independence model was compared to the general loglinear model. If no significant difference was found, the independence model was accepted and the proposed association was thereby rejected. All null hypotheses were rejected at an alpha level of .100. Summaries of the three-way tests for each hypothesis are included below. For the first hypothesis. H1, the "ES" notation indicates that External Support (E) and CBIS Success (S) are associated, while the "E,S" notation implies that External Support and CBIS Success are independent. For all of the rejected models (confirmed associations), the direction of association is also given. The direction of association is determined from the estimates of the loglinear paMIS Quarterly/March 1988 55

Small Business rameters. For each confirmed hypothesis, the results of the relevant two-way classification test are also included as the last test in the list. Hypothesis HI: Greater use of external programming support is positively associated with small business computer-based information system (CBIS) success (ES). The relevant three-way cross-classification tests involve models in which EXTERNAL is independent of SUCCESS. The results of these tests are summarized below. Summary of related three-way tests for external (HQ: E,S): Models Tested TE, TS KE, KS PE, PS IE, iS E, S G' 1.14 6.50 5.08 5.03 0.24 df 46 662 Descriptive

Level .888 .370 .534 .540 .888 Conclusion Accept E, S Accept E, S Accept E, S Accept E, S Accept E, S The models are rejected for a probability value below .100. Conclusion: All five tests indicate that use of external programming is not associated with computer success (E, S). Hypothesis H2: Higher levels of CBIS planning are positively associated with CBIS success (PS). Summary of related three-way tests (Hg: P,S); Conclusion: Chief executive knowledge of computers is associated with computer success (KS). Hypothesis H4: Top management involvement is positively associated with success (IS). Summary of related three-way tests (HQI I,S): Models Tested IK, KS IP, PS

lA, AS IC, CS I, S 27.93 31.87 36.25 30.80 27.89 df 12 12 12 12 4 Descriptive Levei .006 .001 .000 .002 .000 Conclusion Direction Reject I, S + Reject I, S + Reject I, S + Reject I, S + Reject I, S + Conclusion: All five tests confirm that chief executive involvement in computer operations is associated with the success of those computer operations (IS).

Hypothesis H5: Personnel acceptance of computer systems is positively associated with success (AS). Models Tested lA, IS PA, PS RA, RS A, S 13.36 7.63 7.72 3.80 df 12 12 84 Descriptive Levei .343 .813 .461 .434 Conciusion Accept A, S Accept A, S Accept A, S Accept A, S Conclusion: All four tests indicate that employee acceptance of computers is not associated with the success of computer operations

/ A Q\ Hypothesis H6: Greater use of computer conModels Tested TP, TS KP, KS IP, IS AP, AS PC, CS P. S G= 4.56 4.27 8.59 6.90 19.99 3.04 df 8 12 12 12 12 4 Descriptive Levei .803 .978 .737 .864 .067

.551 Conciusion Direction Accept P, S Accept P, S Accept P, S Accept P, S Reject P, S + Accept P, S trois IS Models Tested KC, KS iC, IS PC, PS c, s positivi G^ 8.35 11.02 24.22 4.36 eiy < df 12 12 12 4 associatea Descriptive Levei .757

.528 .019 .359 witn success ^u&). Conciusion Direction Accept C, S Accept C, S Reject C, S + Accept C, S Conclusion: Accept no association between computer planning and computer success excepf in the presence of computer controls (P, S; PCS). Hypothesis H3: Chief executive knowledge of computers is positively associated with CBIS success (KS). Summary of related three-way tests (HQI K,S): Conclusion: The tests indicate that the use of computer controls is not associated with CBIS success exoept in the presence of computer planning (C,S; PCS). Hypothesis H7: Longer use of computers is positively associated with success (GS). Models Tested IK, IS KP, PS KC, CS GK, GS RK, RS K, S

18.38 18.66 18.52 27.25 18.43 12.95 df 12 12 12 12 84 Descriptive Levei .105 .097 .101 .007 .018 .012 Conciusion Direction Accept K, S Reject K, S + Accept K, S Reject K, S + Reject K, S + Reject K, S + Models Tested GK, KS GR, RS

GA, AS Gl, iS GT, TS GE, GS G^ 16.30 10.22 14.81 10.35 11.96 7.75 df 12 8 12 12 8 8 Descriptive Level .1778 .2499 .2522 .5856 .1526 .4582 Conclusion Accept G, S Accept G, S Accept G, S Accept G, S

Accept G, S Accept G, S Conclusion: Length of computer use is not associated with success. 56 MIS Quarterly I Maroh 1988 Small Business Hypothesis H8: More computer training tor employees is positively associated with success (RS). Models Tested RK, KS RA, AS RT, TS GR, GS IR. IS G^ 4.92 7.87 4.20 7.80 7.97 df 6646 6 Descriptive Level .5535 .2481 .3798 .2528

.2400 Conclusion Accept R, S Accept R, S Accept R, S Accept R, S Accept R, S Conclusion: All five tests agree that the level of computer training for employees is not associated with success (R, S). Hypothesis H9: On-site computer use is associated with CBIS success (TS). Models Tested TE, ES TC, CS TR, RS TA. AS GT, GS T, S G^ 7.15 12.56 8.14 10.27 13.70 5.36 df 4 6 46

6 2 Descriptive Level .1283 .0505 .0867 .1139 .0332 .0684 Conclusion Accept T, S Reject T. S Reject T, S Accept T, S Reject T, S Reject T, S Conclusion: On-site computer use is associated with CBIS success (TS). The results of the three-way cross-classification tests for each hypothesis are presented in Table 3. Table 3. Results of Cross-Classification Tests of Hypotheses Hypotheses supported by tests Direction H2 PLAN with SUCCESS (only in presence of CONTROL) + H3 EDPKNOW with SUCCESS + H4 INVOLVE with SUCCESS + H6 CONTROL with SUCCESS (only in presence of PLAN) +

H9 TYPE with SUCCESS N/A Hypotheses not supported by tests HI EXTERNAL with SUCCESS H5 ACCEPT with SUCCESS H7 AGE with SUCCESS H8 TRAINING with SUCCESS Discussion The small manufacturers in this study were found to be making significant use of computers. These firms had been using computers for an average of forty-eight months, with an average of five different computer applications. They spent an average of $4,120 per month on their computer operations, a figure equal to nearly 1 percent of their sales revenues. Twenty percent of the employees were involved with the computer applications, and the chief executives used the computer-generated reports an average of 10 times per month. Thus, the potential impact of these firms' computer operations appears to be great. The primary finding of this study is that the chief executive is the key to the realization of that potential impact. In firms where the chief executive is familiar with computers and is involved in computerization, the computer operations are more successful. External computer expertise is no substitute for chief executive knowledge and involvement because the CEO is the person who understands the factors which are critical to the business' success and the areas where the computer

will have the best payoff. In most cases (73 out of 93), the small manufacturers did solicit outside the company for technical assistance (software development), yet chief executive knowledge and involvement were still significantly related to computer success. In small firms the CEO is the principal information user since he or she assumes responsibility for many of the operational decisions due to lack of managerial staff (Cohn, and Lindberg, 1972); thus the CEO must be involved in decisions as to which systems should be computerized and how they should be computerized. If the small business is to succeed in its computer use, the chief executive must be willing to commit substantial personal energy to the realization of that aim. The association of chief executive knowledge and involvement with success is not affected by the length of computer use. Therefore, chief executive knowledge and involvement are not only important for initial decisions regarding computerization, but also for ongoing computer decisions because computerization is a continuous and evolving process. It is worth noting that chief executive knowledge as measured in this study was a combination of experience with computers and formal training. In the fast growing field of small business computers, continued formal training will be important to computer success. On-site computer use was significantly related to computer success. Apparently, in-house computers

stimulate top management involvement which in turn leads to the application of that resource to higher impact areas. Although on-site MIS Quarterly!March 1988 57 Small Business computers are not necessarily the best way to initiate a firm's computerization, small businesses should develop a plan which specifies a future commitment to acquire computer equipment. With the low cost of small business computer systems, there is a temptation to purchase computers prematurely. This study has shown that low cost computers are no substitute for chief executive knowledge of computers. The practice of planning computer applications and the existence of basic computer controls are important to computer success as well. Small manufacturers should be concerned about implementing and refining computer planning and controls from the moment they begin using computers. Several of the hypothesized associations were not confirmed by the survey results. The length of computer use (AGE) was not associated with CBIS success. A small business should not assume that the successfui use of computers is something that wili come in time with increased computer maturity. Apparently, the mere passage of time is not sufficient to guarantee more successful computer use but rather it is the energies (involvement, planning, and controls) which

are applied to computerization over time that affect CBIS success. Similarly, the level of external support was not associated with success. This result indicates that it is not the source (internal or external) of technical support (software development) which is important but rather how that support is directed toward critical management decision areas through the direct involvement of the CEO. Greater levels of employee acceptance were not associated with higher levels of CBIS success. A minimum level of employee acceptance is a prerequisite for any level of success, but levels of acceptance higher than a threshold value do not necessarily generate higher levels of success. In this study, there were few examples of "poor" employee accepfance and no examples of complete computer failure, so the possible relationship between the lack of employee acceptance and small business computer failure could not be tested. This relationship remains an interesting topic for future research. Only 29 of 93 responding firms engaged in formal training. Formal training alone did not result in greater CBIS success. Presumably, the key computer personnel in the 64 firms without formal training acquired the necessary computer skills before they were hired or through informal on-the-job training such as one employee showing another how to execute a specific fask. Finally, the association between planning and

success was weak since the association was confirmed only in the presence of computer controls. Further analysis suggests that computer planning has its principal impact on success through the involvement of the CEO. Statistical tests revealed that higher levels of planning were associated with higher CEO involvement and that higher CEO involvement led to greater CBIS success. Therefore, the positive impact of computer planning is exerted primarily through the involvement of the CEO in key computer decisions. This is the second major field study of computer success in small manufacturing firms. Raymond (1985) studied computer use in 464 small manufacturing firms in Ouebec and found that in house operations, a greater number of administrative applications, interactive applications, and a higher ranking MIS function were associated with higher levels of user satisfaction and higher levels of system utilization. The present study and the Raymond study involved similar populations and yielded some consistent results where similar variables were tested, as summarized in Table 4. Each study considers variables which the other did not and each had different approaches to measuring small business CBIS success. Future research can build upon these results in order to test the findings. Limitations of the Study

The findings of this study apply only to small manufacturing firms. Whether the results can be extended to small business in other industries is a matter of speculation. For the 93 cases collected, some of the success factor measures did not include enough cases in certain categories. For example, there were too few examples of: 1) significant ongoing employee training (64 out of 93 firms had no ongoing training), 2) in-house program development (16 out of 93 firms), 3) poor or fair employee acceptance (16 our of 93 firms had poor or fair acceptance), and 4) off-site computer services (17 out of 93 firms). The low number of cases in these categories reduced the power of the statistical tests. 58 MIS Quarterly/March 1988 Small Business Table 4. Comparative Findings Average Sample Firm Characteristics Size Age of Computer Use Use of External Support Type (in-house) Associations with Computer Use (SUCCESS) AGE EXTERNAL TYPE This Study 62 employees 48 months

83% 70% This Study NONE NONE YES Raymond 80 employees 72 months 70% 74% Raymond NONE NONE YES The study was not able to determine which of the factors are significant in distinguishing between successful and failing computer operations because there were too few examples of computer failure in the sample. Finally, additional success factors such as software characteristics, hardware alternatives, and vendor relationships could be added to the variables already studied. A single valid and reliable measure of small business computer success should be developed and be applied consistently to all studies in this field. Summary Small manufacturing firms wbich choose to use computers for their information processing requirements need chief executives who are

knowledgeable about computers and who are committed to participating in the strategic decisions concerning computers. These small firms should design and implement computer planning and control systems from the start. Furthermore, they should plan on acquiring their own computer equipment at some point in the computerization process. On tbe otber band, they do not bave to be concerned about developing an inbouse programming staff, since sucb expertise is available externally. References Apcar, L.M. "Angry Computer Users Sue Burroughs for Snags," Walt Street Journal, October 17, 1980, pp. 33,45. Bevis, G. "Managing Resources witb MRP," Small Systems World, 7:2, August 1979, pp. 25-30. Bisbop, Y.M.M., Feinberg, S.E., and Holland, P.W. Discrete Multivariate Analysis: Theory ar)d Practice, The MIT Press, Cambridge, MA, 1975. Bourke, R. "Manufacturing Systems Scene Improving," Small Systems World, 7.2, August 1979, pp. 20-24, 54. Briggs, T.W. "Small Budget, Small Staff? No Big Problem!" Data Management, 18:10, November 1980, pp. 18, 20. Brigham, E.F. and Smitb, K.V. "The Cost of Capital to tbe Small Firm," Ttie Engineering Economist, 13:1, Fall 1967, pp. 1-26.

California t\/lanufacturers Register, Times Mirror Press, Los Angeles, CA, 1981. Cerullo, M.J. "Information Systems Success Factors, Journal of Systems Management, 31:12, December 1980, pp. 10-19. Charleswortb, H.K. "Urgently Required: A Reordering of SBA Priorities to Save Small Business," MSU Business Topics, 20:3, Summer 1972, pp. 45-58. MIS Quarterly I Marcti 1988 59 Small Business Cochran, A.B. "Small Business Mortality Rates: A Review of the Literature," Journal of Small Business Management, 19:4, October 1981, pp. 50-59. Cohn, T. and Lindberg, R.A. How Management Is Different in Small Companies, American Management Association, Inc., New York, 1972. Couger, J.D. and Wergin, L.M. "Systems Management: Small Company MIS," Infosystems, 21:10, October 1974, pp. 30-33. Deeks, J. The Small Firm Qwner-Manager: Entrepreneurial Behavior and Management Practice, Praeger Publishers, New York, 1976. DeLone, W.H. "Firm Size and the Characteristics of Computer Use," MIS Quarterly, 5:4, December 1981, pp. 65-77. Dun and Bradstreet, Inc. The Business Failure Record, Economics Division, 99 Church

Street, New York, 10007, 1981. Ein-Dor, P. and Segev, E. "Organizational Context and the Success of Management Information Systems," Management Science, 24:10, June 1978, pp. 1064-1077. Ein-Dor, P., Segev, E. and Steinfeld, A. "Use of Management Information Systems: An Empirical Study," Proceedings of the Second International Conference on Information Systems, Cambridge, MA, December 1981, pp. 215228. Feinberg, S.E. The Analysis of Cross-Classified Categorical Data, 2nd Ed., The MIT Press, Cambridge, MA, 1980. Garrity, J.T. "Top Management and Computer Profits," Harvard Business Review, 41:4, JulyAugust 1963, pp. 6-12, pp. 172-174. Ginzberg, M.J. "Early Diagnosis of MIS Implementation Failure: Promising Results and Unanswered Ouestions," Management Science, 21 A, April 1981, pp. 459-478. Greenwood, F. "The Ten Commandments of Small Business Computerization," Journal of Small Business Management, 19:2, April 1981, pp. 61-67. Hamilton, J.S. and Chervany, N.L. "Evaluating Information System Effectiveness Part I: Comparing Evaluation Approaches," MIS Quarterly, 5:3, September 1981, pp. 55-69. Hanes, L.F. and Ramage, W.W. "Productivity Measurement for Computing and Information

Systems," Proceedings of the Ninth Annual Society for Management Information Systems Conference, 1977, pp. 81-86. Hansen, J.R. "How Vendor Strategies Affect Users," Small Systems World, 7:11, May 1980, pp. 24-27. Heise, J. "Personnel Acceptance, Management Understanding Are Success Factors, Says Small Computer Turnkey," Data Management, 18:10, 1980, pp. 26-29. Huysmans, J.H.B.M. The Implementation of Qperations Research, John Wiley and Sons, Inc., New York, NY, 1970. King, W.R. and Rodriquez, J.I. "Evaluating Management Information Systems," MIS Quarterly, 2:3, September 1978, pp. 43-51. Klatt, L.A. Small Business Management: Essentials of Entrepreneurship, Wadsworth Publishing Co., Belmont, CA, 1973. Kriebel, C.H. "Evaluating the Quality of Information Systems," in Design and Implementation of Computer-based Information Systems, Szyperski and Grochia (eds.), Sijthoff and Noordhoff, Germantown, MD, 1979, pp. 29-43. Lucas, H.C., Jr. "Performance and the Use of Information Systems," Management Science, 21:8, April 1975, pp. 908-919. Lucas, H.C., Jr. "Empirical Evidence for a Descriptive Model of Implementation," MIS Quarterly, 2:2, June 1978, pp. 27-41. Markland, R.E. "Can Computers Help the Small

Businessman?" Journal of Systems Management, 23:4, April 1972, pp. 27-30. Markland, R.E. "The Role of the Computer in Small Business Management," Journal of Small Business Management, 12:1, January 1974, pp. 21-26. McKinsey and Company. "Unlocking the Computer's Profit Potential," McKinsey Quarterly, Fall 1968, pp. 17-31. Neidleman, L.D. "Computer Usage by Small and Medium Sized European Firms: An Empirical Study," Information & Management, 2:2, May 1979, pp.67-77. Newpeck, F.F. and Hallbauer, R.C. "Some Advice for the Small Business Considering Computer Acquisition," Journal of Small Business Management, 19:3, July 1981, pp. 1723. Raymond, L. "Organizational Characteristics and MIS Success in the Context of Small Business," MIS Quarterly, 9:1, March 1985, pp. 37-52. Raysman, R. "Manager Involvement Needed in Computer Selection," Harvard Business Review, 59:5, September-October 1981, pp. 5458. Rockart, J.F. "Chief Executives Define Their Own Data Needs," Harvard Business Review, 57:2, March-April 1979, pp. 81-93. 60 MIS Quarterly/March 1988 Small Business

Rotch, W. Management of Small Enterprises: Cases and Readings, University of Virginia Press, Charlottesville, VA, 1967. Schollhammer, H. and Kuriloff, A.H. Entrepreneurship and Small Business Management, John Wiley and Sons, Inc., New York, 1979. Seibt, D. "User and Specialist Evaluation of System Development," in Design and Implementation of Computer-based Information Systems, Szyperski and Grochia (Eds.), Sijthoff and Nordhoff, Germantown, MD, 1979, pp. 29-43. Senn, J.A. and Gibson, V.R. "Risks of Investment in Microcomputers for Small Business Management," Journal of Small Business Management, 19:3, July 1981, pp. 24-32. University of California, Los Angeles, Health Sciences Department. BMDP-79, University of California Press, Berkeley, CA, 1979. U.S. Small Business Administration. Facts About Small Business and the U.S. Small Business Administration, Washington, D.C., 1981. U.S. Small Business Administration. Annual Report FY 1983, 1, Washington, D.C., 1983. VanLommel, E. and DeBrabander, B. "The Organization of Electronic Data Processing (EDP) Activities and Computer Use," Journal of Business, 48:3, July 1975, pp. 391-410. Vazsonyi, A. "The Rise and Decline of a Mainframe: A Progress Report," Interfaces, 11:1,

February 1981, pp. 31-35. Walker, E.W. "Investment and Capital Structure Decision Making in Small Business," in The Dynamic Small Firm: Selected Readings, E.W. Walker (Ed.), Austin Press, Austin, TX, 1975. Weber, H. and Tiemeyer, E. "Teaching Information Systems to Small Business Management, Information & Management, 4:6, December 1981, pp. 297-303. Weike, R.J. and Konsynski, B.R. "An Examination of the Interaction Between Technology, Methodology and Information Systems: A Tripartite View," Proceedings of the First International Conference on Information Systems, Philadelphia, PA, 1980, pp. 32-48. Williams, R. "Considerations for Computer Implementation in a Small Business," Parts 1-4, Interface Age, May, June, July, and August, 1978. Zmud, R.W. "Individual Differences and MIS Success: A Review of the Empirical Literature," Management Science, 25:10, October 1979, pp. 966-979. About the Author William H. DeLone is Assistant Professor of Management and Information Systems at the Kogod College of Business Administration, The American University in Washington, D.C. He received his Ph.D. in Computers and Information Systems from UCLA. His current research interests

include the use of computers in small business and the measurement of MIS effectiveness. MIS Quarterly/March 1988 61

Vous aimerez peut-être aussi