To begin with, the researcher will conduct an extensive analysis on secondary sources such as internet, books and academic journals with the aim of finding the possible factors that are likely to determine performance of a company, and then use the insight to design the questionnaire. Based on the result of the questionnaire, the factors that seem to carry the heaviest weight will be selected and used as the dependent and independent variables for statistical analysis. In this regards, change in sales since 2000 and change in demand have been selected as the dependent variables. The predictor variables will include innovativeness, specialization, lasting customers relations and high-quality marketing strategies. Nowadays, many companies have adopted very advanced technology, which keeps changing day in day out and, therefore, it becomes very important to uphold innovativeness all through. Any company that is not innovative is liable to be thrown out of market by those, which are innovative. This is the reason innovativeness seems to cause considerable impact on performance of many firms (Carter et al., 1994). Furthermore Dyer (1997) affirms that a company can increase its productivity through specialization, which leads to adoption of a particular niche based on its core business. Taylor (1997) has cited customer relations as the most important factor helping firms to become successful. This is because customer is considered the most important party in a business, and if they are treated well they keep coming back. Marketing strategy is meant to publicize the products of a company, and if a company is very committed to it and designs high-quality marketing plans, then the customers is attracted to purchase products from the company hence boosting sales (Wynarczyk et al., 1993). Conceptual Model (CM) Company 1 Company 2 3.0 Regression Analysis 3.1 Linear Additive Model for entire sector PF1 = [X1, X2, X3, X4] Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 .940a .884 .866 .766 .884 47.711 4 25 .000 a. Predictors: (Constant), X4, X1, X2, X3 ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 112.114 4 28.028 47.711 .000b Residual 14.686 25 .587 Total 126.800 29 a. Dependent Variable: PF1 b. Predictors: (Constant), X4, X1, X2, X3 Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .516 .871 .593 .558 X1 -.128 .094 -.118 -1.362 .185 X2 .625 .133 .468 4.705 .000 X3 .321 .111 .493 2.900 .008 X4 -.012 .090 -.021 -.131 .897 When PF1 (Change in sales since 2000) is compared with the four indecent variables, the above regression and correlation results are found. The adjusted R squared is very high at 86.6%, which indicates that the model is not subject to a lot of errors. Only 13.4% of the prediction is attributable to error and hence the model is quite reliable. The p-value, which is indicated in the table as Sig. is less than 0.005 implying that there is enough evidence that the dependent variable (PF1), has some statistical relationships with at least one of the four predictors. Amongst the four predictor, X2 has the strongest prediction power as indicated by the Sig. figure, which is less than 0.001. PF2 = [X1, X2, X3, X4