A sample of 61 venture capital-backed firms was drawn from a population of the 76 venture-capital-backed U.S. firms that had an IPO in 1991 and were less than five years old at the time of the IPO. Fifteen firms were eliminated because no prospectus from 1991 was available from the Securities and Exchange Commission or because the 1993 Compact Disclosure database did not include them. The size of the firms ranged from zero sales to $660 million in 1991, with a mean of $79 million and a median of $13 million. The number of employees ranged from 18 to 47,000 with a mean of 1,668 and a median of 171. For additional description of this database and for an extended discussion of the operationalization procedures followed for this study, the interested reader is referred to Bloodgood et al. (1996).
Criterion and Predictor Variables
Inbound Logistics; Manufacturing Operations; Outbound Logistics; Marketing & Sales; Service: Each of these five criterion variables were coded using a content analysis procedure, using information provided in the IPO prospectus of each of the 61 ventures. In each case the variable was binary coded with a value 0 if that particular activity was not internationalized and 1 if it was internationalized. To enable us to test the first hypothesis, we summed the values for the three individual upstream categories and divided by three to arrive at the percent of upstream activities that were at least partially internationalized. We followed a similar procedure of summing the values for the two individual downstream categories and dividing by two, to arrive at the percent of downstream activities that were at least partially internationalized.
International Work Experience: The number of directors with previous international work experience was determined from the Management section of the IPO prospectus. The range for this variable was from zero to four.
Cost advantage; Product differentiation; Market differentiation: Using a content analysis procedure, each of these three variables was binary coded using information provided in the IPO prospectus. In each case, the variable was coded, respectively, a 1 or 0 to signify whether or not the venture used it as a source of competitive advantage.
Innovation: This variable was measured using information from the income statement in the IPO prospectus as R&D expenses expressed as a percentage of total expenses. The range for innovation was from zero percent to 100 percent, with a median of 11 percent.
Size: This variable was expressed in terms of the number of employees of the venture. Since this distribution was skewed (18 to 47,000 employees), the logarithmic form was used. The median number of employees was 171.
To the extent that financing a venture may impact its ability to internationalize, we controlled for the extent of leverage at the time of IPO for each firm. To minimize variation in internationalization due solely to industry level phenomena, we controlled for industry growth rate and industry profitability.
Leverage: This variable was computed as a ratio of total liabilities over total equity using information provided in the venture's IPO prospectus. The range of this ratio was from -226.4 to 103.0, with a median of .27.
Industry growth rate and Industry profitability: The industry growth rate was computed using average firm sales in the industry for the period 1989-1991, expressed as a percentage of 1989 sales. Industry growth rate ranged from -27% to 110% with a median value of 30%. Industry profitability was computed for 1989 as the ratio of average firm gross profit to average firm sales. Industry profitability ranged from 25% to 69% with a median value of 40%. The information used to compute these ratios was taken from Dunn's Industry Ratios based on the same 4 digit SIC code as the individual firm information.
Table 1 presents the summary statistics for and zero order correlations among all the variables. The first hypothesis was tested using a paired sample t-test. The rest of the hypotheses were tested using the binomial logistic regression (Logit model) procedure since the dependent variable being examined was categorical. We also provide a simple frequency description of the characteristics of those ventures that internationalized at least some part of their value chain activity. Table 2 presents the results of the paired sample t-test for Hypothesis 1. Table 3 presents the results of the logistic regression analyses for testing Hypotheses 2 through Hypotheses 7.
An analyses of the 61 ventures revealed that 28 ventures had internationalized some part of their primary value chain activities. Of these, 47 percent had internationalized at least one of their activities, 7 percent had internationalized two activities, 14 percent had internationalized three activities, and 32 percent had internationalized all five of the activities. Among these INVs, 11 percent had internationalized only their upstream activities, 39 percent had internationalized only their downstream activities, and 50 percent had internationalized both their upstream and downstream activities.
Hypothesis 1, which predicted that downstream activities in the INV are more likely to be internationalized than upstream activities, was supported (p-value of paired sample t-test < .01). The second set of hypotheses (2a, 2b, 2c), which predicted that international work experience was positively related to the internationalization of upstream activities, was also supported. The third set of hypotheses (3a-3c), which predicted that cost advantage would be positively related to the internationalization of upstream activities, was not supported. The fourth set of hypotheses (4a-4e), which predicted that product differentiation is positively related to the internationalization of both upstream and downstream activities, was not supported. The fifth set of hypotheses (5a-5e) which predicted that market differentiation is negatively related to the internationalization of both upstream and downstream activities, was not supported. The sixth set of hypotheses (6a-6e) which predicted that the innovativeness of the INV will be positively related to the internationalization of both upstream and downstream activities, was not supported. The seventh set of hypotheses (7a-7e), which predicted that the size of the INV is positively related to the internationalization of both upstream and downstream activities, was supported.
This study examined the primary value-chain activities most likely to be internationalized by a sample of venture capital-backed U.S. firms at the point of their initial public offering. The study also examined the relationship between initial conditions and internationalization of these value-chain activities. We found that the extent to which U.S. based high potential ventures internationalized their upstream activities differed from the extent to which they internationalized their downstream activities. We also observed that the decision to internationalize upstream activities was dependent on the presence in the venture of executives with international work experience. Further, we observed that venture size was a significant determinant of the likelihood that the venture would internationalize its value-chain activities.
Our results suggest that HPVs are likely to internationalize their downstream activities to a greater extent than they are their upstream activities. They also indicate that HPVs which internationalize their upstream activities had executives with international work experience in their top management team. Whether this relationship is a reflection of better international networks established through experience or of superior knowledge of international markets cannot be ascertained from our data. Our data showed that large-sized HPVs were more likely to internationalize their primary value-chain activities than their smaller counterparts. The effect of size could imply that the larger venture either had more resources to pursue a broader scope of operations or it had access to a greater wealth of knowledge of international experience among its officers (Bloodgood et al., 1996). However, this relationship was weaker for Marketing & Sales (p < .1) when compared to the other four activities. A possible inference of this finding is that, regardless of its size, Marketing & Sales is among the first activities to be internationalized by the venture if it desires to enter the international arena. This inference is consistent with our observation that of the ventures that had entered the international arena, 89 percent had internationalized their Marketing & Sales activity.
The adoption of a cost advantage strategy was not related to internationalization for upstream activities. Also, the market differentiation strategy was unrelated to the internationalization of both upstream and downstream activities. The lack of a relationship between cost advantage and the internationalization of upstream activities may be attributable to lack of adequate variance on this variable in the sample. Of the 28 firms that had internationalized at least one of their value chain activities, only 3 had pursued a cost advantage strategy. A similar argument can also be made for the lack of a relationship between market differentiation strategy and the internationalization of both upstream and downstream activities. Of the 28 firms that internationalized at least one of their value chain activities, only 2 had opted to pursue a market differentiation strategy.
Product differentiation was unrelated to the internationalization of all the value-chain activities, except for Marketing & Sales. Product differentiation was negatively associated with the decision to internationalize the Marketing & Sales activity, the opposite of what we had hypothesized. We are unsure as to the inferences that can be made here and instead only speculate about this finding. This result when examined in conjunction with the relationship observed between innovation and Marketing & Sales implies that ventures that engage in innovation or pursue product differentiation as a means of competitive advantage have a lower likelihood of internationalizing their Marketing & Sales activity. One explanation could be that firms that pursue product differentiation are likely to engage in innovation (significant correlation coeff. in Table 1, p <.05) and hence their resources are being invested in research and development (innovation) type of activities, as opposed to in internationalizing the Marketing & Sales activity. The relationship between innovation and Outbound Logistics was also significant and negative, contrary to our hypothesis. Again, a speculative inference would suggest that ventures engaging in innovation have not yet established a market presence and hence do not engage in any outbound logistical type of activity.
Our results hold several implications for entrepreneurs and others who are associated with entrepreneurial efforts. First, the results suggest that while entrepreneurs in the U.S. may adopt a strategy of rapid globalization while entering the international markets, they need not internationalize, at the onset, all the activities in their value-chain. This suggestion is consistent with Porter's (1986) argument that when the international pattern of competition tends to be multidomestic, the internationalization of primarily downstream activities is vital to competitive advantage. Second, for a venture that seeks to internationalize its upstream activities, our results suggest that it should include officers with international work experience in its top management team. Third, ventures that intend to internationalize all their value-chain activities would be more successful if they start big.
Two primary limitations of this study merit comment. First, the sample included only U.S. ventures and hence the findings cannot be generalized to the internationalization approach used by new ventures outside the U.S. Nonetheless, since U.S. firms are often under less pressure to internationalize their activities rapidly, studying U.S. firms is a more stringent test of the process used for internationalizing the value-chain. Second, ours is a cross-industry study and it is likely that the rate of internationalization is driven by industry conditions. Although we controlled for some of the industry effects, we were unable to control for all industry-specific variance. However, our strong findings, despite this limitation, suggest the power of the model that we tested.
As the phenomenon of entrepreneurial ventures that are internationalized at their inception becomes increasingly common, future research in this area can benefit from both theoretical advancements related to INVs and methodological alternatives to our approach. Porter (1986) argued that the specific value-chain activity that should be internationalized is dependent on whether industry characteristics are multidomestic or global. The extent of internationalization varies across industries which may impact a new venture's proclivity towards internationalization. Identification of other firm-level resources and their interactions with the competitive weapons chosen by the venture should shed further light on this process. We have also not examined the performance implications of internationalizing the different value-chain activities. Methodological alternatives could involve investigations of a longitudinal nature, studies in other countries and cultures, and also single industry studies.
To conclude, our study shows that, consistent with Oviatt and McDougall's (1994) arguments, U.S. firms need not necessarily adopt the conventional wisdom of internationalizing their activities in a sequence of relatively predictable stages. Our study is consistent with Bloodgood et al.'s (1996) conclusion that "the circumstances which foster internationalization may be more commonly faced by new U.S. ventures than was the case in the past...."
Previous Page | Main Menu | Next Page
© 1997 Babson College All Rights Reserved
Last Updated 1/15/97 by Geoff Goldman & Dennis Valencia
To sign-up for the Center for Entrepreneurial
Studies' publication lists,
please register with the Entrepreneurship WebTeam.