Building on the authors' previous work on the transformation of nascent entrepreneurs' ideas into actual new firms, this paper goes the next step and examines what factors may be associated with survival and varying patterns of sales growth in the early years of a firm's existence. New firms do not start at the same level of sales. They do not all grow at the same rate. Nor do they necessarily grow at all or grow with the same consistency over time. Some new firms start with high, initial-year sales and continue to soar while others start high and slide backwards. Some firms start more modestly but grow rapidly and consistently while others start modestly and continue modestly growing and shrinking sporadically. Many entrepreneurs seek to learn if there are certain steps which they can take at the outset to better assure that they will both survive and thrive. The charge of this paper is to reveal what can be learned from the experiences of several hundred new firms that might better inform the entrepreneur.

Several others (Keeley and Knapp, 1993; Keeley and Knapp, 1994; Reynolds, 1993) have explored this question in some form. Reynolds examined the characteristics of the top two percent of a sample of 1424 younger companies to see if they exhibited a number of systematic differences from the other firms in the sample. He found that in some cases they did, such as the larger, starting-team size. Keeley and Knapp have tried to differentiate among the fastest growers, small starters, and venture-capital backed firms to see, within a small sample, if there are systematic differences among the firms based on where they started. They found that the differences between high performers and small start-ups was not easily accounted for: in fact, personal qualities of the entrepreneurs, the way they learn and adapt, the quality of the original opportunity, or some combination are thought to contribute to the differences in growth patterns, not the more easily measured characteristics of the team members or the business.

Our study attempts to build on the previous work but sets some stricter parameters, such as using firms of the same age and using a larger sample than that used by Keeley and Knapp. We also have expanded the definition of growth. Rates of growth are just one component. Consistency of growth is another. The starting point - first years sales - is a third. Growth rates are not unambiguous: smaller firms are likely to have faster relative growth rates than larger firms. That is why we examine the sales level at which firms start as well as the speed with which they move. Furthermore, another concern is whether a firm can grow each and every year, not just sporadically. Consistent growth may lead to higher levels of sales and to fewer headaches.

We initially examine the issue of survival and whether any of the characteristics we measured are related to a new firm's ability to survive. We then go on to examine the relationship between numerous characteristics of start-up teams and firms and the fastest-growth firms. Third, we differentiate among three components of growth: level of start, rate of growth, and consistency of growth. We seek to determine whether any of a long list of start-up team and business factors is associated with survival and each growth pattern. This latter exercise helps us to understand the role of definition in shaping some of our conclusions regarding what it is that contributes to "sales growth."

The one issue on which we ultimately focus is the potential role of the public sector to assist firms in reaching the different growth patterns. Specifically, what role did use of business assistance programs make in how the firms developed? We found in a previous analysis (White and Reynolds, 1994) that business assistance programs did seem to play an important role in the transformation of nascent entrepreneurs into fledgling new firm owners. The second part of that story is whether the use of these programs in the early stages of the firms' lives is associated with survival for at least four years and what growth paths, if any, are also associated with greater use of this assistance. If utilization of these programs is associated with greater success, it is yet another reason to further support these business assistance efforts.


The information used in this analysis is taken from two surveys of new firms in Wisconsin. Baseline information was gathered from a random sample of firms that began business no earlier than January 1, 1987. The sample was drawn from the Unemployment Compensation files of the state. This means that these were firms that had at least one employee and the resources to report their presence to the state. In other words, the sample was drawn from a slightly select subset of all firms: it contained those firms which were viable enough to have employees, which had the resources to pay unemployment compensation insurance, and which were willing to comply with the reporting laws. Data were gathered from 449 firms in 1992-93, and follow-up survey information was collected in late 1994 and early 1995 from 332 of the original respondents. The data from the two surveys were combined to form a longitudinal record for each firm.

Most of the information gathered was straight forward. But problems were created by our survey instruments in the interpretation of the meaning of such indicators as "first year sales." If a firm reported its start date as July 1990 and its first year sales as $50,000, should we interpret the "first year" period to span from July 1990 to July 1991 or July 1990 through December 1990? In order to create what we construed to be a more accurate measure, we did try to create an estimated full year of sales figure for those firms that did not begin business in January of their first year. When this computed figure was substantially larger than the second-year sales figure, we used the initially reported figure in lieu of the estimated figure, assuming that the firm had reported an annual figure. Thus, our first-year sales figures are a combination of reported and estimated sales. That combination appears to be the best estimate possible of the actual sales which did occur.

We also had to interpret the lack of responses to our attempts to undertake follow-up interviews. For some firms it was clear that they had died. Their phone numbers were disconnected, and no references could be found to them. For others their survival is not as clear cut. One group of firms could be reached by telephone but never had time to be re-interviewed. These we counted as still alive. Common members of this group were doctors' offices. A second group had telephone answering machines that listed their firm name. We left messages but never talked to any individuals. It appears that these are marginal businesses, at best. If any of these reported employment to the state in 1994, they were counted as a survivor. Thus, a firm was alive in 1994-95 if we could interview it, talk with principals in it but not interview it, or locate it by phone but not have any discussion with the principals, if there were some other proof it still existed.

A third issue is how many years of firm life do we need in order to differentiate fast- from slow-growing firms. We decided that we would examine those firms which were able to last at least four years. That is, they report employment or sales for at least four years. Some of these may not have been alive in 1994, but they had had at least four years of business activity by the time of their initial interview or they continued to report employment to the state but were not around at the time of the follow-up interviews. Since very few of our sample first reported sales or first employment in 1992, the size of our sample was hardly diminished by using a four year definition.


Entrepreneurs go through stages of development. The period of incubation as founders attempt to put together a business we have termed "nascent entrepreneurship." Some portion of these nascent entrepreneurs then successfully transform into new business owners or what we call "fledgling firms". Over a period of years these fledglings transform into "established businesses". This analysis examines that third phase, the transformation over time of fledgling firms into established businesses. Before examining growth patterns, therefore, a more fundamental question is what contributes to survival, the successful transformation from fledgling to established firm. Are there characteristics of the start-ups which are associated with greater chances of survival? In other words, are there some steps entrepreneurs can take that might increase their chances of surviving their first four years? We address that briefly.

It may well be that few of the more easily measurable qualities of a start-up team or the firm can be associated with survival. One might get that impression based on some previous studies (Gartner, Mitchell, and Vesper, 1989; Keeley and Knapp, 1993; Keeley and Knapp, 1994; Reynolds and Miller, 1990; Woo, Cooper, and Dunkelberg, 1991). Nevertheless, to be consistent we have examined qualities of both the start-up team and the firm to see what, if anything, is associated with survival.

Among our sample of firms, all of which had reported having employees at some point, some 66% lived through their fourth year. This is in sharp contrast to 88% which survived three years. The higher early survival may be attributable to the fact that this is not a sample of all firms but a sample of all firms which had and declared they had employees early in their firm lives. In a few years we can extend the analysis to see whether any relationships change as firms age. For our analysis here we will examine only those firms which had lived at least four years.

In terms of start-up team characteristics there were few differences between survivors and decedents. Team sizes, gender representation on the start-up team, experience of start-up team members with previous start-ups, college completion of team members, and the like were extremely similar for both groups. Similar percentages of both sets of firms employed one or two persons their first year. But there were significant differences as well. Age of the entrepreneur matters: entrepreneurs under 30 years of age were three times more common among the decedents than the survivors. Surprisingly, those with experience starting other new businesses, those who took more entrepreneurship or business courses, and those with higher incomes in 1994 were more likely to fail. Otherwise the start-up teams were quite similar.

Business characteristics were also rather similar. The amount of informal investment, first year assets, bank loans, and team financial commitment all differed, but none did so significantly. The same can be said of 14 of the 15 other characteristics which we tracked. Even though decedents had 50% higher first year sales, the difference is not statistically significant. The only significant difference that emerged is that those firms which survived were more positive about the quality of the assistance they received from the various business assistance programs. This seems logical, as those who survived should be more positive about what turned out to be sound advice. The short of it is that there are few characteristics which we measured which can help to predict whether a new firm will survive its first four years.

Top of Page
Previous Page | Main Page | Next Page

1997 Babson College All Rights Reserved
Last Updated 4/1/97 by Cheryl Ann Lopez

To sign-up for the Center for Entrepreneurial Studies' publication lists,
please register with the
Entrepreneurship WebTeam.