Capitalism continually renews itself with new business ventures fulfilling the vital function of creating hundreds of thousands of new jobs (Birch, 1979; Birley, 1986; Reynolds, 1987). When semiconductor giants such as Intel and Advanced Micro Devices declined in product quality and market share relative to Japanese firms in the mid-1980s, for instance, Silicon Valley continued to prosper as dozens of emerging businesses generated new jobs and growth rates exceeding 40 percent (Powell, 1992). Emerging ventures fill voids occasioned by the demise of larger corporations, but they enhance long-term economic vitality only when they survive and thrive.

The contribution of new businesses to our economy makes the performance of emerging firms an important focus for entrepreneurship research. Low and MacMillan (1988: p. 142) describe most entrepreneurship studies as contributing in some way to "explaining and facilitating the role of new enterprise in furthering economic progress." Researchers model new venture performance (McDougall, Robinson & DeNisi, 1990), predict success, (Roure & Keeley, 1990), describe developmental processes of emerging businesses (Singer, 1995) and document patterns of failure (Bruno, McQuarrie & Torgrimson, 1992). Indeed, a large body of research has emerged seeking to identify the correlates and antecedents of new venture performance.

The lack of reliable, valid, and meaningful performance measures, however, hampers researchers’ efforts (Brush & Vanderwerf, 1992; Chandler & Hanks, 1993). Theorists criticize existing measures, lamenting the use of borrowed methods and measures which do not deftly fit "disjointed, discontinuous, and non-linear processes" of emerging businesses (Bygrave, 1989). They appeal to researchers to use concepts, measures, and methods grounded in theory and knowledge of entrepreneurial phenomena (Low & MacMillan, 1988). Meanwhile, empiricists bemoan the paucity of objective accounting and market performance measures used to examine larger cor-porations. Abundant criticisms and problems exist, but researchers offer few new solutions.

We view the development of reliable, valid, and meaningful performance measures as imperative to our efforts to "explain and facilitate" new venture performance. These measures should measure current financial performance, but also capture dynamic dimensions of the health and vitality of new business ventures. Thus, the purpose of this paper involves proposing a new way of thinking about the performance of emerging companies, exploring the patterns of growth in emerging businesses rather than relying on just static measures of sales level or summary statistic measures such as average growth. Thus, our study responds to calls for researchers to adopt "a more contextual and process-oriented focus" (Low & MacMillan, 1988) concerning one of the most central construct in entrepreneurship research, new venture performance.



Prior studies prompt researchers to evaluate at least two dimensions of new venture performance: (1) shareholder value or equivalent measures of earnings in closely-held firms, and (2) growth. (Chandler & Hanks, 1993; Robinson, Kunkel & Hofer, 1994). In analyzing a variety of new venture performance measures, Robinson et al. (1994) showed that shareholder value exhibited the strongest ability to effectively corroborate previously established linkages between new venture strategy and the venture’s economic performance. Unfortunately, market data only exist for publicly held firms and not for the vast majority of startups that are not traded publicly.

In cases where archival data sources do not exist, researchers require a proxy measure of the value firms generate. Robinson et al. (1994) showed using non-parametric statistical techniques that shareholder value yielded a strong relationship with net income, as in Ball and Brown (1968), and exhibited strong relationships with ROS, ROA, ROIC, and ROE--corroborating results by Buzzell and Gale (1987).. These findings suggest that in closely-held companies--those not publicly trading stock and for which shareholder value measures do not exist--earnings measures of performance identified by Robinson et al. (1994), and Chandler and Hanks (1993) should represent meaningful proxies of shareholder value generated by the firm.

Researchers must, however, question whether they can rely on subjective measures reported by founders who may intentionally or unintentionally bias data.. Researchers have used both subjective (e.g. Sapienza, 1992; Covin & Slevin, 1990; Chrisman & Leslie, 1989) and objective performance measures (e.g. Chandler & Hanks, 1994; Duchesneau & Gartner, 1990; Sapienza, 1992). Furthermore, Chandler and Hanks (1993) have shown objective data on size (sales level) and growth (sales growth) display relatively reliable and valid attributes; they described self-reported earnings as reliable when measured in objectively defined categories. Brush and Vanderwerf (1992) similarly provided evidence for the reliability and validity of selected self report performance measures. The available evidence supports the reliability and validity of self-report measures, with some exceptions such as return on investment, return on assets, or return on equity (Chandler & Hanks, 1991).

Researchers have tended to rely on stock and rate dimensions of growth performance, sales level and its first difference across years. Chandler and Hanks (1993) describe size (measured by sales level) and sales growth as relatively independent. They provide evidence of the reliability of self-reported sales growth measures when reported for the past year. Self-reported growth measures, however, become increasingly unreliable with the passage of time (Chandler & Hanks, 1993) and appear incapable of capturing the dynamic nature of emerging businesses.

We believe extant growth measures suffer from two major weaknesses. First, emerging businesses oftentimes do not exhibit monotonic sales growth, so single-year sales or growth (first difference) figures may capture aberrations and not represent the true health of the organization. Conversely, if researchers use growth averages, such summary statistics do not capture complex growth patterns across time and may not accurately represent the firm's current performance.

In this research we seek to better understand growth by analyzing not only growth rates, but growth patterns. In our review of the new venture performance literature, we found no studies that analyzed the patterns of sales growth. Yet, researchers may gain substantial insights by analyzing the events accompanying various patterns of growth. They can examine growth rates, the volatility of change, plus gleaning insights about the peaks and valleys of performance and the antecedent and accompanying events. Indeed, if researchers respond to call for grounding performance measures in theory and knowledge of entrepreneurial phenomena, or if they adopt more contextual and process orientations, researchers must examine patterns in growth.

Proposition 1: Researchers can glean insights into factors related to new venture growth by analyzing patterns of growth, antecedents, and concurrent events.

Second, when researchers use sales growth as an indicator of performance, they implicitly assume that faster growth is desirable. They hold that fast-growth indicates better performance than slow-growth in empirical analyses, despite normative treatments of startup warning founders of the perils of growing too fast (Timmons, 1990). Excessive growth strains human and physical resources, and demands ever increasing cash flows. A shortages of human resources increases confusion and the potential of mistakes, while the absence of sufficient cash flows cause rapidly growing firms to burn themselves out in a relatively brief flash. Abundant anecdotal evidence suggests that companies sometimes go out of business because they have done too much, too fast--even though they generate value. Given this scenario, more moderate growth rates may be more desirable--a possibility linear models of venture performance have not captured. The plausibility of the assumption of non-linearity can be tested simply by attempting to fit curvilinear models to growth data. If excessively rapid growth leads to subsequent company burnout, we would expect a quadratic relationship between growth rates at one period of time, and growth rates in a subsequent time period.

Proposition 2: A curvilinear relationship exists between a firm’s growth rate at one period of time, and growth rates measured in subsequent time periods.

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.