Strategies and Methods for Supplier Selections - Strategic Sourcing of Software at Ericsson Mobile Platforms Caroline Raning & Johanna Vallhagen February 2007 Department of Industrial Management and Logistics, Faculty of Engineering, Lund University, SE-221 00 Lund, Sweden The software industry is today one of the most rapidly growing branches of business. In many companies, software has increasingly replaced hardware as a great part of the total value delivered to customers. Despite this, methods for evaluating software suppliers are a domain that has not been paid much attention to date, although the importance of selecting the right supplier is likely to be even more important in the future. This article is based on a master thesis study, which was conducted at the Sourcing Department of Ericsson Mobile Platforms in Lund. The aim of the article is to elucidate how a supplier selection process should be defined to support the sourcing of software items from software suppliers, and what kind of evaluation criteria that need to be considered to make a correct decision. Key words: Supplier Selection Process, Mobile Platforms, Software Sourcing, Hardware Sourcing, Self Assessment Template, Supplier Classification Introduction Since suppliers are increasingly contributing to the competitiveness and overall success of many companies, supplier selection is regarded one of the most important decision making processes. Numerous researches and studies have been conducted within the field of supplier evaluations, but they have all almost exclusively focused on the evaluation of suppliers in the manufacturing industry. Little attempt has been made to identify which aspects that need to be considered when evaluating non-manufacturing companies such as software providers even though it is likely to assume that there are fundamental differences between these businesses. In spite of the common belief that the software business lack industry standards and management guidelines, and suffers from bugs, late deliveries, and increasing costs, it seems as the benefits achieved by evaluating software suppliers has not yet been realised. [1] The Sourcing Department at Ericsson Mobile Platforms (EMP) is indeed familiar with the problem previously described. As a leading platform supplier of major mobile technologies, EMP is dealing with a broad range of suppliers, offering hardware as well as software, but no formal process supporting the selection of suppliers exists within the business unit today. The main task of this study was to evolve strategies and methods for supplier selections, applicable to the complex mixture of hardware and software items in the business environment of EMP. A considerable part of the study was to elucidate the differences between hardware and software and determine how to manage hardware and software suppliers respectively. This article will however only focus on the findings related to software sourcing. Problem description The lack of guidelines and supporting materials for supplier selections at EMP currently result in a subjective judgement of each supplier, based on individual competences rather than the common strategy of the company. The purpose of the master thesis study was thus to create a supplier selection framework, adapted to the specific business environment of EMP. The concrete areas of investigation were as follows: How should a supplier selection process be defined to be applicable at EMP? According to which criteria should software suppliers be evaluated? Method The study was conducted according to the systems approach. The open system defined for the study comprises the Sourcing Department of EMP and its supplier relations in particular, but also other departments within EMP and Ericsson. During the study, the authors physically participated in the system, which enabled an abductive research method
and thus continuous altering of theoretical and empirical studies. Empirical data was mainly obtained through interviews, and validated through repeated work-shops. Theoretical frame of reference The literature study conducted for the thesis was broad, and the theoretical frame of reference thus covers several fields of research. For instance, the theoretical framework encompasses Supplier evaluation methods presented by van Weele [3] and Axelsson & Wynstra [2], and Purchasing portfolio models such as the ones suggested by Kraljic [4], Bensaou [5], and Olsen & Ellram [6]. The theoretical framework also includes a summary of Assessment criteria suggested by various researchers such as Çebi & Bayraktar [7] and Pearson & Elram [8], although they are mostly applicable to hardware components and manufacturing suppliers. Moreover, with an attempt to elucidate the characteristics of the software industry, software is compared with hardware and services respectively. To name a few, papers written by Allen & Chandrashekar [9], Ellram & Billington [10], and Ulkuniemi & Pekkarinen [1] have been studied. Current supplier selection process Supplier evaluations at EMP are currently defined as the activity of assessing the supplier s responses from the RFI/RFQs. As depicted in Figure 1, illustrating the current supplier selection process, there is a great lack of established documents, routines, and guidelines supporting the different steps of the process. Moreover, there is no common system for documenting and filing the results from the different steps for future sourcing activities. As a consequence, Sourcing Officers have developed their own way of working, inhibiting the deployment of a common sourcing strategy. Supportive documents Receive sourcing assignment Prepare RFI/RFQ (adapt predefined template) Send out RFI/RFQ Evaluation of incoming responses from the RFI/RFQ Select supplier Figure 1 Current supplier selection process The only supportive document that exists today is a standardized RFI/RFQ template, including technical, business, legal, and quality requirements. However, the document is working more as rough guidelines and does not consider the differences in purchasing hardware and software, neither the differences between sourcing objects within these two categories. Since these kinds of item specific questions are not included in the standardized RFI/RFQ, the Sourcing Officers have to add, remove or change the questions in the template in order to adapt it to the specific circumstances of the different sourcing situations. Accordingly the RFI/RFQs that are sent out to the suppliers are time consuming to create and also differ considerably from case to case. New suggested supplier selection process In order to create a common strategy for how to work internally, as well as towards suppliers, we have suggested the supplier selection process illustrated in Figure 2 to be implemented at the Sourcing Department. Supportive documents Receive sourcing assignment Classify sourcing object Prepare RFI/ RFQ Weight assessment criteria Send out RFI/ RFQ Evaluate answers from RFI/RFQ Select supplier for further negot. Figure 2 New supplier selection process Filing The fundamental difference between our suggested process and the routines currently used at EMP is that we have divided the existing supplier selection activities into subprocesses and created some supportive documents in order to facilitate the different steps. As illustrated above, the supplier selection process is initiated by the reception of a sourcing assignment, describing the overall scope of the sourcing case. Subsequently, a classification of the various sourcing objects should be accomplished in order to make use of differentiated strategies towards suppliers. For each and every product class, a commodity specific self assessment template has been developed with the purpose to render the preparation phase of the RFI/RFQ more efficient. Guidelines for how the overall assessment categories should be weighted have also been established. The underlying reason is to minimize subjective judgements and assure that the evaluation of the supplier is reflected by the company s strategy. When it comes to the weights associated with the specific assessment criteria, we realize that it might be
difficult to have predefined, standardized weights that fit each and every sourcing object within a product class. The weights should however be determined before the RFI/RFQ is sent out to the supplier in order to make sure that the supplier is judged objectively. To facilitate the subsequent step where the Sourcing Officer should evaluate the incoming responses from the RFI/RFQ, guidelines associated with the self assessment templates have been developed. The last, but probably one of the most important steps in the supplier selection process is to document and file the results from the different sub-processes for future sourcing activities. Since the Sourcing Department is currently facing a high employee turnover, we argue that an implementation of this step is crucial to make the transitional period between two employees run more smoothly. Not only a system for internal feed-back should be developed, but also a process for how to give feed-back to suppliers. The fact that these activities are currently not performed is somewhat remarkable. We argue that the deployment of this new process is likely to result in that the supplier selection activities will get faster to accomplish. The increased support will make it easier to understand each others working tasks, and consequently easier to cover for each other. As a result, the Sourcing Department will become more flexible as an organization. It should moreover be stressed that the initiation of a structured supplier selection process will lay ground for future improvements and strengthened relationships with the suppliers. Because EMP will be forced to formulate explicit specifications in an early stage, the suppliers will have great potential to fulfil EMP s requirements. Software assessment criteria As previously described, the new suggested supplier selection process implies that a classification of the sourcing objects should be accomplished in order to exploit differentiated strategies for various items. However, within the Sourcing Department of EMP, there is a strong reluctance towards segmenting software sourcing objects. Although general differences can be distinguished between the various software products, Sourcing Officers argue that the reality is fragmented implying that availability, importance, and risk exposure are unique for every single item to be purchased. Therefore, Sourcing Officers claim that sourcing strategies must be developed from case to case instead of being shared by a group of sourcing objects. As a consequence, despite many efforts to classify the software items, this study has not resulted in a definite classification proposal. With regard to the commodity specific self assessment templates, this implies that only one template has been developed for software. Hence, to identify relevant assessment criteria for software items, the authors have been forced to have a rather broad perspective to cover all different aspects that might be associated with different types of software items. To explore what criteria that need to be considered when evaluating a software supplier, extensive theoretical studied have been altered with empirical research such as interviews with the Sourcing Officers at EMP. During the interviews it was revealed that software suppliers are currently most often selected based on their abilities to deliver an adequate technical solution, to a reasonable price, under sufficient legal arrangements. However, both theoretical and empirical findings suggest that these areas should be extended to cover a broader spectrum of assessment categories. Therefore, as depicted in Figure 3 below, the commodity specific self assessment templates developed for this study comprise 7 different categories; Technology, Quality, Logistics and Supply, Legal, Company Profile, Business, and Finance. COMPANY PROFILE BUSINESS LEGAL FINANCE SUPPLIER LOGISTICS & SUPPLY TECHNOLOGY Figure 3 Assessment Topics QUALITY In the following section, we intend to briefly elucidate the different aspects, under each and every category, that have been identified as
having a significant importance when evaluating software suppliers. Technology Technical requirements are prominent on the items sourced at EMP since technological excellence is one of the key competitive advantages of the company. Because software components often require customization and adaptation before being implemented in the platform, EMP presents a Statement of Work for the software supplier, which basically illustrates the characteristics of the platform and the surrounding software and hardware to describe how the software must be developed or modified in order to fit into the platform environment. This implies that the software development process is important to evaluate in order to eventually receive an accurate product on time. Although the Statement of Work contains the most important technical requirements needed to secure the desired functionality, we argue that some additional questions should be added to further assure that the product of inquiry is prioritized by the supplier and that the supplier will be a potential partner also in the future. For instance, it is very interesting to know the current roll out plans of coming releases in order to find out how fast the development is expected to proceed. Man years required to develop the product is also of significance since it is a good indicator of code complexity, and hence the estimated value of the software. Quality Although the sourcing department at EMP has experienced shortcomings of the quality of externally developed software in the past, there are currently no routines of assuring sufficient quality of software sourced from third parties. Although EMP is presently not focusing on software quality, it is a common belief among the Sourcing Officers that future demands will force EMP to think differently. To manage short product life cycles and increased competition, there is no time for dealing with insufficient quality. A considerable problem at EMP is however that the competence required to assess software quality is not prevalent within the company today. Not knowing how to assess software quality accordingly implies that the competence required to develop accurate self assessment questions and communicate requirements to the supplier is lacking. Our recommendation is therefore to allocate resources to this area in the future. Until then, we believe that the Sourcing Officers should stress the importance of the issue and ask the supplier some general quality questions, exploring whether the supplier is striving towards creating or maintaining a functioning process across the software development unit. Such general questions should include whether the supplier has a quality management system implemented, if the software development process is well defined, documented and divided in sub-processes with clear milestones. It should also be investigated whether the supplier has a defined testing process and whether all requirements are tested by accurate test scenarios. Logistics & Supply According to Sourcing Officers, estimating development lead time is extremely hard when developing software. Deliveries behind schedule are indeed a problem identified both by EMP and by other actors in the telecommunication business. Despite this, it seems as these late deliveries are generally accepted, and that companies do not really work proactively with these issues. Therefore, we claim that EMP should try to make suppliers become better at approximating lead times and put monetary penalties into operation when delivering too late. Legal Since software sourcing objects are most often not purchased, but licensed, legal aspects are crucial to assess in order to determine obligations and rights connected to each business case. Normally, EMP distributes a Software License Agreement to the supplier in the negotiation phase, but we argue that it might be advantageously to include some parts of the agreement already at an early stage of the sourcing process. The following legal aspects should be considered: License grant Warranty Limitation of liability Indemnification Governing law and choice of jurisdiction It should be stressed that minor deviations can be agreed on during negotiations, but if the supplier is completely unwilling to accept
EMP s legal requirements, it is very unlikely that EMP will continue with the negotiation process. In addition with the fundamental legal aspects mentioned above, we emphasize the importance of covering also the issue of free open source software, which is increasingly implemented in software components. If not treated correctly, these can be devastating for EMP. It should therefore be investigated whether the supplier has all rights and interests required to forward the software, and under what premises free open source software is implemented in the software components offered by the supplier. Company Profile At present, no questions treating more longterm aspects of a business relationship are included in the existing RFI/RFQ template. We argue that these kinds of aspects should not be neglected and therefore emphasize the use of criteria aiming at exposing the strategic plans, market position, and future growth and success of the supplier. The overall purpose when analyzing the answers of these questions must be finding a company that is in alignment with the EMP strategy of the specific sourcing item. Business When EMP licenses software, there are two different business models frequently used. Either the software is sourced with an upfront payment, or with continuous royalty payments. The total estimated price is what determines which business model that is preferred from case to case. In order to compare prices and settle on a favourable business case, EMP wants to know whether the supplier is compliant with any or both of the suggested business models, as well as the estimated cost structure and price. In order to secure that no unexpected costs associated with the business will appear, EMP moreover wants suppliers to specify additional costs before starting the negotiations. Suppliers often try to spread out incomes on various issues such as maintenance and support fees, development costs etc, whereas EMP wants it to be covered by the initial fee or the price per unit. Finance Many companies in the software industry are start-up companies with very few employees, but can nevertheless deliver superior software components. Assuring functionality, EMP is therefore enforced to license software from very small companies, despite the risks connected with these businesses. In such cases, we argue EMP should pay some attention to evaluating the company s financial status. EMP must make sure that the supplier has the required financial stability to handle unexpected events such as patent disputes. Without financial muscles, warranties and indemnifications are useless in case of a legal dispute with another company. In order to assure financial stability, a number of key ratios should be examined and compared with similar companies within the software industry Discussion Both our theoretical findings and our study at EMP reveal that the software business in many ways is immature. The major issue, as we see it, is that this seems to be generally accepted within EMP. It is a common believe that it is nearly impossible to anticipate software development time and what problems that might arise along the way. Accordingly, there is a resistance to create a structure of how to handle these issues, and an industrialization of the software business just seems very distant. As described previously in this article, processes related to software sourcing at EMP are few and rather cursory, and Sourcing Officers have difficulties in specifying for instance quality and delivery requirements to suppliers, which consequently results in late deliveries and inferior software code. Many Sourcing Officers claim that it is the intangible and unique nature of software components that makes it so difficult to establish a formal process for managing software suppliers. However, the new suggested supplier selection process indicates that the establishment of a common way of working does not have to be that impossible or difficult. Also, the proposal of a new self assessment template with a much broader spectrum of assessment criteria indicates that there are definitely opportunities to concretize the intangible characteristics of software, and start communicating expectations and requirements to suppliers in a more structured way. It should be stressed that we do realize that this industrialization of the software business will not be easy to carry out. We definitely agree
that software is complex, but we also think that there is a point in asking oneself why it is complex. It seems in many cases that there are no sound arguments; people are just claiming that it is complex because it has always been that way. Hence, although it is not obvious for companies within the software business to develop and control processes in order to manage the software supply chain, we argue that companies at least should try to think on these lines. Since the importance of software are most likely to increase in the future, we argue that companies will experience great problems if they do not manage to break down the large cloud of activities related to software sourcing into more controllable processes - processes that can easily be spread within the company, as well as to suppliers and customers. Conclusions and Recommendations Returning to the areas of investigation initially formulated, some conclusions can now be presented. Primarily, a supplier selection process has been suggested with the main purpose of striving for consistency, both in terms of the work conducted internally, as well as towards suppliers. By adopting differentiated strategies for various sourcing objects, and developing supportive documents and guidelines related to the different supplier selection activities, the overall process will be more consistent with the company s strategy, and also get faster to accomplish. The suggested process aims to accentuate the importance of utilizing various competences and resources within the company in order to reach synergy effects. The second area of investigation, treating the criteria according to which the software supplier should be evaluated, revealed that software suppliers are currently selected according to their technical abilities, price offerings, and legal terms and conditions. We argue that these areas are not sufficient to create a true picture of the supplier and its offerings, and have therefore extended the existing evaluation template to also include categories such as finance, quality, company profile, and logistics and supply. Since the software market is rapidly changing and consist of a variety of suppliers, we emphasize a greater focus on assessing the company and its present, as well as future strategies. Furthermore, we argue that EMP should require more from their software suppliers in terms of quality assurance and delivery accuracy. By acquiring competence in software quality management, EMP could develop better requirement specifications and require considerably more in terms of product quality. Similarly, the importance of on time deliveries must be further emphasized in order for software suppliers to follow EMP demands. Instead of accepting that software development is a complex task, Sourcing Officers should find and make use of strong incentives to improve performance and change the habitual behaviour of the software industry. It should finally be stressed that to date, very little research has been conducted on software supply chains in general and software sourcing in particular. The software industry has grown rapidly over the years as a result of many products, such as cars and mobile phones, containing more and more software. Since software is nowadays often what differentiate and add value to products, it has become a more important part of the total product offering. Thus, from an academic point of view, or suggestion for future studies is to further investigate the similarities and differences between software supply chains and traditional ones, in order to find out which theories and models are valid for software supply chains and where supplementary studies are necessary. References [1] Ulkuniemi, P, and Pekkarinen, S, (2005) Managing competitive software component supplier relationships Journal of Purchasing & Supply Management 11 (2005) 97 106 [2] Axelsson, B, and Wynstra, F, (2002) Buying Business Services John Wiley & Sons Ltd, United Kingdom [3] van Weele, A J, (2002) Purchasing and Supply Chain Management, Third edition, Thomson Learning [4] Kraljic, P, (1983) Purchasing must become supply management, Harvard Business Review, 83509, 110-117 [5] Schary, P B, and Skjøtt-Larsen, T, (2003), Managing the Global Supply Chain, Second edition, Copenhagen Business School Press [6] Olsen, R F, and Ellram, L M, (1997) A Portfolio Approach to Supplier Relationships, Industrial Marketing Management 26 101-113
[7] Çebi, F and Bayraktar, D, (2003) An integrated approach for supplier selection Logistics Information Management, Vol 16, No 6, p. 395-400 [8] Ellram, L M, and Pearson, J N, (1995) Supplier selection and evaluation in small versus large electronics firms, Journal of Small Business Management, Oct 1995; 33, 4, p.53-65 [9] Allen, S, and Chandrashekar, A, (2000) Outsourcing Services: The Contract Is Just the Beginning Business Horizon, March-April 2000, p. 25-34 [10] Ellram, L M, Tate, W L, Billington, C, (2004) Understanding and Managing the Services Supply Chain The Journal of Supply Chain Management, Fall 2004; 40, 4, p. 17-32