127 / 274
臺大管理論叢
第
27
卷第
3
期
127
We introduced Models 9 and 10 to assess the mediating effect of exploitation on prior
exploration and performance. According to Baron and Kenny (1986), exploitation can be
considered a mediator if (a) prior exploration significantly predicts firm performance, (b)
prior exploration significantly predicts exploitation, and (c) exploitation significantly
predicts firm performance controlling for prior exploration. In Table 5, Model 7 shows that
prior exploration has a positive effect on a firm’s performance (
p
< .10). In Table 4, the
coefficients in Models 4 and 5 support a positive prior exploration effect on exploitation (
p
<
.01). Furthermore, in Table 5, Models 9 and 10 suggest that exploitation significantly and
positively impacts a firm’s performance with prior exploration controlled (
p
< .01). When
exploitation is included in the prior exploration regression on firm performance with
controls, the influence of prior exploration is insignificant; indicating a full mediation effect.
Thus, the above results significantly support Hypothesis 4, that exploitation mediates the
effect of prior exploration on performance.
5. Discussions, Limitations, and Conclusions
Balancing exploration and exploitation is an important strategy for firms to outperform
others, yet not much empirical research clearly elucidates how a firm can do so to gain
benefits. We posed two research questions for our study: (a) how does uncertainty influence
exploration and exploitation, and (b) what was the relationship between exploration and
exploitation and its effect on firm performance? Drawing on real options reasoning, we
found that uncertainty is positively related to exploration and that exploitation was a
mediator that fully mediated the relationship between prior exploration and firm
performance.
This study finds that uncertainty is positively related to exploration while it poses an
insignificant effect on exploitation. The empirical results suggest that firms invest more
exploratory investments in the face of uncertainty, indirectly implying a higher value for
options created by such investments under uncertainty.
Our findings suggest that exploitation plays a key role in materializes the outcomes of
exploration and that exploration and exploitation are balanced over time to create value for
firms. This result supports an alternative enhanced view (Vermeulen and Barkema, 2001;
Lavie and Rosenkopf, 2006) that balances exploration and exploitation over time and over
domains. Additionally, the result on the positive effect of prior exploration on exploitation
highlights the role of path-dependency in technology management, echoing Cohen and
Levinthal’s (1994) statement that “fortune favors the prepared”.