SUPLIER SCORING AND ITS IMPACT ON PROCUREMENTS

Size: px
Start display at page:

Download "SUPLIER SCORING AND ITS IMPACT ON PROCUREMENTS"

Transcription

1 SUPLIER SCORING AND ITS IMPACT ON PROCUREMENTS Autor: Larisa Gavrila 1 Introduction The main purpose of this paper is to present a logical and measurable way of fundamenting a decision based on statistical procedure. In order to reduce the subjective part of any decisional process, I have looked into a statistical method that approaches that topic. I have investigated the way Analytical Hierarchy Process handles decision takers at management level. Several studies have focused on this subject and after analyzing them I was able to implement a template, constructed based on principles provided by the Analytical Hierarchy Process in order to prioritize a set of criterias. Moving forward, from the result provided by AHP method, I have decided to test the cost reduction possibility on a series of companies. The cost reduction possibility was one of the main outputs provided by the AHP method. In order to test the efficiency I have used the Data Envelopment Analysis. The DEA analysis besides outlining the efficient and inefficient companies, has also shown where certain policies and internal procedures should be revised and to what extent they should incorporate best practices from efficient companies. Now from DEA analysis, we know which companies should bring improvements to certain inputs. By applying the least square method in Eviews I was able to measure the impact of inputs variance on outputs. 2 Analytical Hierarchy Process Method Analytical Hierarchy Process (AHP) was developed by Saaty in 1980 (Saaty, 1980) as a method of analysis of decisions based on a hierarchy of decision components. As noted Leleur in 2004 (Leleur, 2004) proved to be one of the most applied methods AHP and mentioned in most textbooks and guidelines on AHP. This method is essentially an interactive one in which a

2 decision maker or a group of decision makers and analysts transmit preferences can be debated or discussed opinions and results. The method derives largely from theories about human behavior, including those relating to the process of thinking, logic, intuition, experience and learning theory. AHP also developed a linear additive model, but its format standard procedures used to derive the weights and scores achieved by alternatives which are based, respectively, paired comparisons of criteria and between options. Therefore, the AHP is based on the construction of a series of arrays "paired comparisons". The aim of this approach is that the ranking or weighting of each criterion which describes the importance of the contribution of each of these criteria to the overall objective. If the criteria are broken down into a number of sub-criteria comparisons pair is repeated for each one of the levels of this hierarchy (Saaty, Decision making with the analytic hierarchy process, 2008). One of the great advantages of AHP is the calculation of the inconsistency as a ratio between the decision maker inconsistency and randomly generated index. This index is important for decision makers because they ensure that their judgments are consistent and that the final decision is taken well. Inconsistency index should be less than Although a higher index of inconsistency requires reassessment comparisons in pairs, decisions obtained in certain cases it may also be considered as the best alternative. 3 Data Envelopment Analysis Method This approach focuses on the variations in performance between companies. The Data Envelopment Analysis makes no assumption about the functional form: it is a non-parametric approach to performance evaluation. With envelopment analysis, the benchmark against which the relative performance of companies can be measured is the efficient frontier. Given a sample of companies, all companies should be able to operate at a level of peak efficiency, which is determined by the effective companies in the sample. These efficient companies are generally called "peer companies" and determine the efficiency frontier. Companies that define the efficiency frontier using a minimum quantity of inputs to produce the same amount of production. The distance from the efficient frontier is a measure of the efficiency or lack thereof.

3 In DEA, the indication of inputs ("variables influenced by the company") and outputs ("variables not directly influenced by the company") is required. To determine the effectiveness of distributors values, we used a form of DEA minimizing inputs. Benefits of DEA: The main advantage of this method is its ability to take into account a multiplicity of inputs and outputs. It is also useful because it takes into account the returns to scale in the calculation of efficiency, incorporating the notion of efficiency by increasing or decreasing the size and production levels. Disadvantages of DEA: the results are potentially sensitive to the selection of inputs and outputs, so that their relative importance must be analyzed prior to the calculation. However, there is no way to verify whether these results are relevant. The number of successful companies at the border tends to increase with the number of input variables and productions. When there is no relationship between the explanatory factors (in inputs and / or outputs), the DEA considers each company as unique and fully effective and efficient notes are very close to 1, the method loses then its power of discrimination. 4 Empirical studies on AHP and DEA usage The AHP method has been used before within supplier selection rationale. In the paper Supplier Selection Using Analytic Hierarchy Process: An Application From Turkey, AHP methodology is used to determine the best supplier for purchasing computer and printers in General Directorate of Land Registry. In order to buy 2000 computers and 500 printers for General Directorate of Land Registry, the writers Betül Özkan, Hüseyin Başlıgil, Nergis Şahin decided to determine the best supplier for this purchase so they have proposed a set of criterias as it is described bellow (Özkan, Başlıgil, & Şahin, 2011) : The main criteria are; General and organizational structure of the firm, Production capability, Service quality and Price. Their sub-criteria are detailed below. General and organizational structure of the firm - Employees number and quality: the number of the well-qualified employees should be high, because the products are distributed to different areas all over the country. - Sector experience: it would be better that the potential supplier firm have worked with different public concerns like General Directorate of Land Registry before.

4 - References: it shows the customer satisfaction degree of the supplier while working in similar sectors. - Communication capability: having strong communication while solving the problem and between other suppliers, it occurs a trust. And it is expected having a strong communication between customer an the firm in marketing and handling process. - Service adequacy: the supplier should have a service infrastructure that is integrated and has a good knowledge and experience and can solve the problems quickly. - Capital: it is preferred that the supplier is in good economic condition. The firm can choose the supplier with a big capital while purchasing some complex products. Production capability - Delivery date appropriateness: the supplier should deliver the products at the expected quality and at the time that is specified in the arrangement. - Material appropriateness: the materials that will be used in production should have in good quality and some standards. - Technological knowledge: it is expected that the supplier can meet customers requirements using the new technology. - Material lead time: it shows the power of the communication between the domestic and oversea suppliers. The lead time should be short. Service quality - Packaging and carrying capability: the material should be delivered to the desired place at the right time perfectly. - Flexibility: it shows the quick response to the changes related to the material. - Sale and service network: it is expected that the material can be supplied from the nearest location in- and after the agreement time. The firm has to have common authorized service network. -Customer satisfaction: the product quality, the use of product, solving the problem and the relationship between the customers are all related to the customer satisfaction. - Research and development activities: these are activities that are done for improving the product quality. Price

5 This is one of the most important criteria. Generally the cheapest one will be preferred. But the cheapest one is not always the best. Using these criterias, a questionnaire is being created and sent to sourcing experts, the surveys are evaluated using Saaty s 1-9 scale, every expert answered the survey individually and then the geometrical average was calculated and a single value was found. After the pairwise comparison, the weights are calculated and the best supplier is selected. Data Envelopment Analysis has been used within the supply chain. Weber and Desai (1996) compared 6 suppliers of one of the Fortune 500 companies. They have identified inefficient suppliers for the purpose of negotiation leverage. They have also underlined how supplier s performance need improvement to increase the overall efficiency. Narasimhan (2001) proposed a framework for supplier performance evaluation and rationalization, combining supplier performance and efficiency scores. Talluri, Narasimhan and Nair (2006) applied a case study in a division of Fortune 500 pharmaceutical company. They compared the CCDEA (chance-constrained) results to deterministic DEA results and highlighted its usefulness in the decision-making process. 5 Taking a decision using AHP method. The sourcing department of a telecommunication company has been asked to answer a questionnaire over a certain business case scenario. The business case scenario is presented bellow : By being in the position of a sourcing manager you will need to identify a hardware supplier for our new Alpha project. Within this project we will need a long term relationship with the purchasing supplier and also the supplier will need to engage to a cost reduction plan in order to assure the proper prices in the future. We are interested in buying on yearly basis more than 1 million chips that will be integrated in telecommunication systems. The new system integration will affect all our customers. The selection criteria will be focusing on quality, cost reduction, price, financial stability.. Three sourcing managers have been asked to fill in a questionnaire that has been transposed into an excel macro for AHP method. The first input information was inserted in the excel file : - number of criteria, scale ( 1 = standard linear scale 1 to 9; 2 = logarithmic ;3 = Square root ; 4 = Inverse linear ; 5 = Balanced ; 6 = Power ; 7 = Geometric ),

6 - number of participants, - alpha (threshold for acceptance of inconsistency ), - selected participant (for more than 1 participant you can select whose participant s result to be displayed, participants are numbered from 1 to 10 according the input sheets for pair-wise comparisons, when selecting 0, the consolidated result for all participants will be shown, using the geometric mean of all decision matrices); - objective (text) to describe the project/category, - author (text, optional), - date (date, optional) - The consolidated result is the most important one because it combines all three participants answers. The results demonstrate that the most important criteria is cost reduction, followed by quality and financial stability, the output is presented below : If the value of Consistency Ratio is smaller or equal to 10%, the inconsistency is acceptable. If the Consistency Ratio is greater than 10%, we need to revise the subjective judgment. In this case the value is 3.6% and so we can conclude that the subjective evaluation over supplier preference is consistent. 6 DEA efficiency analysis

7 The data was split in three main clusters : companies with a small total inventory, companies with a medium total inventory and companies with a high total inventory. Each cluster was analyzed separately and inefficient suppliers have been identified. Also, by using DEA method I was able to mention to what extent each inefficient supplier should adopt policies and best practices from efficient suppliers. Below is the demonstration for the first cluster. First cluster analysis : in this cluster I have included companies with a reduced total inventory, bellow are stated the companies along with their inputs and output. Company name Research & Development Total Inventory Accounts Payable Long Term Debt Income Before Tax Integrated Device Juniper Logitech Nvidia Broadcom AMD Sandisk Northrop Seagate Mototola When calculating efficiency scores, we can conclude that within this cluster we have four companies that are not efficient : Broadcom, AMD, Sandisk, Motorola. In order to be efficient Broadcom has to reduce its inputs with ( 40.53%), AMD with (87,48%), Scandisk (25.54%) and Motorola with (16.97%). Efficiency scores Integrated Device

8 Juniper Logitech Nvidia Broadcom AMD Scandisk Northrop Seagate Motorola By putting this data into a graph, it will look like bellow and we will be able to distinguish even faster which company is closer to the efficient ratio 1 :

9 Also by analyzing the efficient peers and weights we can see to what extent the inefficient companies should adopt best practices from efficient companies.

10 Efficient peers and weights Integrated Device Juniper Logitech Nvidia Northrop Seagate Integrated Device Juniper Logitech Nvidia Broadcom AMD Scandisk Northrop Seagate Motorola For example, the objective for the inefficient company Broadcom is to become, roughly speaking, 21% of Northrop company, 4% of Nvidia and 75% of Integrated Device. This is useful information, especially when some coefficient is close to 1 as in this case. The management at Broadcom has to adopt methods and practices from Integrated Device. Looking at this excel export and analyzing the numbers I can conclude that in all the cases inefficient companies should guide their attention to Integrated Device company s best practices, so in this particular case Integrated Device company would act like a leader. 7 The impact on output when reducing inputs Eviews study case Now that we know who should reduce their inputs in order to be efficient we can start analyzing to what extent the output will change when decreasing certain inputs. In order to obtain this result, I have imported in Eviews the data from all the companies analyzed and by using least square method I have obtained the bellow output.

11 I have reached the conclusion that the total inventory should be reduced in order to improve the income before tax, all other inputs were in positive connection with the output. This also proves that the cluster separation by total inventory was a good approach from the beginning. 8 Conclusions Within this paper I have demonstrated a way of taking a decision in a less subjective manner. By using the AHP method, the selection criteria can be obtained after questioning the concerned responsible persons and by taking in consideration everybody s opinion. Depending on the level of importance offered to each criteria by each questioned person, a combined weight is calculated and after that the hierarchy is built. Within my study case I have chosen the criteria quality, cost reduction, price and financial stability. After building the hierarchy the criteria cost reduction has been pointed out as the most important one. Just like in a real business case situation, a set of suppliers have been chosen for further RFx ( request for information, quotation, price). Those suppliers were selected for the efficiency test with DEA method. The suppliers were separated in three main clusters based on total inventory, each cluster was analyzed separately and inefficient suppliers have been identified.

12 Also, by using DEA method I was able to mention to what extent each inefficient supplier should adopt policies and best practices from efficient suppliers. After analyzing efficient and inefficient companies, I have used the least square method in the econometric tool Eviews in order to determine how the inputs are influencing outputs so that I could underline what kind of strategy each inefficient company should fallow. I have reached the conclusion that the total inventory should be reduced in order to improve the income before tax, all other inputs were in positive connection with the out put. This also proves that the cluster separation by total inventory was a good approach from the beginning. It would be very interesting to see for future researches a different combination between the method AHP and DEA. I have used AHP for building the selection criteria hierarchy, but it can also be used to directly select suppliers. If combined the weights resulted from the AHP method with the weights resulted from DEA method, we could probably be able to develop even a stronger decision model. 9 Bibliography Andrei, T. (2008). Introducere in econometrie utilizand Eviews. Bucuresti: Editura Economica. Bratianu, C. (2000). Management strategic. Bucuresti: Editura Ceres. Certo, S. (2002). Management modern. Bucuresti: Editura Teora. Emilian, R. (2003). Management. Bucuresti: Editura ASE. Farell, M. (1957). The Measurement of Productive Efficiency. Journal of the Royal Statistical Society. Filip, G. (2005). Decizia asitata de calculator - decizii, decidenti, metode de baza si instrumente informatice asociate. Bucuresti: Editura Technica. Gheorghita, M. (2010). Econometrie Avansata. Bucuresti: Editura ASE. Holsapple, C. W., & Whinston, A. B. (2000). Decision.Support Systems. Minneapolis: West Publishing Company. Leleur, S. (2004). Systemic Planning. Polyteknisk Forlag: TU Denmark. Özkan, B., Başlıgil, H., & Şahin, N. (2011). Supplier Selection Using Analytic Hierarchy: An Application from Turkey. London: Proceedings of the World Congress on Engineering. Power, D. (2000). A Decision Support Systems Glossary. DSS Resources COM. Saaty, T. (1980). The Analytic Hierarchy Process. New York: McGraw Hill.

13 Saaty, T. (2008). Decision making with the analytic hierarchy process. Pittsburgh: University of Pittsburgh. Saaty, T., & Vargas, L. G. (2006). Decision Making with the Analytic Network Process: Economic, Political, Social and Technological Applications with Benefits, Opportunities, Costs and Risks. New York: Springer. Verboncu, I. (1999). Ghid metodologic pentru manageri. Bucuresti: Editura Technica.