One crucial matter in routine life is decision-making. Although decision-making in everyday life appears to be easily dealt with by one-on-one comparisons, the same is not true with decision-making in business. In business management decisions, problems are more complicated than those in individual life because most business situations fall into the category of multi-criteria decision-making (MCDM) problems. Therefore, appropriate multi-criteria decision methods should be carefully selected to solve these problems. This study uses the analytic network process (ANP), one of the widely available multi-criteria decision methods. However, this popular method generally ignores the uncertainty inherent in the input data. Therefore, this paper proposes an improved process that considers uncertainty by using Monte Carlo analysis with input values then applied to the ANP procedures. This proposed method is implemented by a real business that produces roof tiles; the primary goal of the study is to select among newly developed roof formulas by considering the uncertainty and interrelation among decision criteria and elements as well as alternatives. The outcomes of study accurately rank the new product formulas. Furthermore, the results of improved method differ the rankings produced by the original ANP. The observed dissimilarities mainly result from uncertainty consideration discussed in this study. be independent from other elements, but, in fact, most real-life problems involve interaction as well as dependency among elements (Saaty, 2001). Thus, an improved method, the analytic network process (ANP), was developed to overcome this problem. The ANP is extended to consider both dependency and feedback. For these reasons, the ANP was later consolidated with several quality methods to produce an improved and more comprehensive decision-making method (Kahraman et al., 2006). However, ANP approximates the prioritization of the competing candidates. Eigenvalues from ANP are calculated from interrelations among input data computations and are normally applied as a basis of prioritization; however, they still lack inclusion of the uncertainty and probability information and lead to inappropriate decision-making (Wey, 2008;Hsu and Pan, 2009).Normally, there is available research that considers the uncertainty in single input data such as the calculation of cost or quality; nevertheless, there has been little attention focused on the uncertainty between cost and quality coordination in MCDM problems. Moreover, most of ANP studies largely ignored the uncertainty (Huang and Tzeng, 2007). Generally, MCDM anticipates the costs, opportunity, benefits, resource consumption, material supplies and other factors, but the exact values of these various attributes are difficult to estimate. Thus, decision-makers may be operating with incomplete knowledge as to the complete data values. Consequently, uncertainty must be included in the data. Some researchers have proposed using the integrated MCDM with an uncertainty module. Lee and Kim (200...