Optimising your Supplier Portfolio
This article originally appeared in Outsource Magazine Issue #29 Autumn 2012
In today’s global services outsourcing arena, increasing numbers of companies adopt “multisourcing”; that is, they select and combine information technology (IT) and business services from multiple providers. The decrease in deal size, well-documented by the Global TPI Index, is one indicator that buyers are dividing their business among multiple providers. Gartner – which has promoted this term since the publication of the 2006 book Multisourcing: Moving Beyond Outsourcing to Achieve Growth and Agility by its consultants Cohen and Young – has been continuously emphasising multisourcing as one of the top trends in the sourcing of IT and BPO services.
There are, however, well-known risks involved in increasing the number of providers. Every sourcing manager knows that more vendors mean more headaches: more coordination issues, more searches, and more contracting costs. Yet these are only surface costs; a much more serious problem with increasing the number of suppliers within a single function – a practice commonly engaged by today’s buyers – is that the buyer may lose economies of scale within this function, thereby increasing production costs. These increased costs can be reflected immediately in higher bidding rates or creep in over time as vendors are unable to optimise the function they were charged with. For example, a large global bank outsourced its account reconciliation business to three different vendors getting what seemed to be a low charge rate from each due to competitive pressures. None of the vendors, however, had the right incentives to invest in a new information system necessary to optimise the function. Had the three contracts been combined, the investment in the system would’ve paid off within the lifetime of the contract.
Single sourcing, however, may not be the answer either as the trend to multisource has arisen in response to problems associated with lock-in, an inability to tap into best-of-breed capabilities, and increased operational risks. A common wisdom for resolving the tension between too few and too many providers is to focus on a handful of strategic partners so as to address these risks while avoiding full lock-in. This rule of thumb may not be sufficient for generating the greatest benefits from today’s complex sourcing opportunities. In some situations a single vendor may be the best solution whereas in others, numerous vendors may deliver the best results. Moreover, associating more vendors directly with lower lock-in is a bit naïve. In some cases, increasing the number of vendors may reduce lock-in when such vendors are easily substitutable for each other. In other cases, however, vendors may have diverse capabilities and each vendor may have invested in learning specific clients’ needs making them less-than-ideal substitutes. For example, multisourcing is very common in IT application development. This knowledge work, however, is often done at the level of teams who invest in learning specific business needs and technological constraints of the client organisation. Having multiple vendors involved in this work hardly reduces lock-in.
Our multi-year research project focussed on understanding the multisourcing practices of diverse clients and their impact on providers and business outcomes; we find it useful to think about vendor portfolios, which include not only the breadth (how many vendors), but also the depth (how co-invested the parties are) of vendor relationships. This view allows us to discuss situations in which multiple vendors can substitute for each other (relationships are of low depth) versus not (each relationship is quite deep). The main thing to recognise is that large breadth does not automatically equal small depth in sourcing relationships (see Figure 1). Sourcing relationships should and do allow for polygamy (many deep relationships) as well as for uncommitted monogamy (one shallow relationship). By recognising that depth and breadth are two different dimensions of a multisourcing strategy, clients can start analysing which portfolio structure they want for each function. To do so, they need to consider how varying the breadth and depth of their vendor portfolio will result in different performance outcomes.
First, let us consider critical drivers of outsourcing success. As illustrated in Figure 1, every sourcing relationship can be judged based on how well it reduces the costs and risks involved in performing a given function, while increasing the output (service quality, speed, flexibility, and innovation). One can assume that whatever aspects of outputs can be measured, they were put in a contract and thus have specific associated costs and risks involved. Thus, what we care about in a sourcing strategy beyond known costs and risks is how to get the vendors to produce things that are not in the contract (we refer to them as “intangibles”).
What drives these costs, risks, and intangibles? For any given function, we would argue that a combination of basic production factors (vendor’s generic capability, vendor’s client-specific capability, and economies of scale within a function) as well as the client’s ability to reduce strategic and operational risks by reducing vendor dependency (switching costs) shape the outcomes of the relationship.
Now, consider how the choice of the number of vendors and the depth of each relationship impact these drivers. For a given outsourced function, developing and maintaining a broad vendor portfolio allows the client firm to access a diverse set of capabilities, both generic and specific to the client. These superior capabilities may bring cost advantages to the client as vendors are able to deliver at lower costs. Thus for example, in the famous ABN-AMRO case of multisourcing to IBM, Accenture, TCS, Infosys, and Patni in 2005, each vendor brought capabilities to the table that allowed it to lower production costs. While this case is easy to make for multisourcing across function (e.g. IBM handling infrastructure vs. TCS handling application maintenance), even within the function of each provider (e.g. application development), each one of the five preferred suppliers can have some prior experience that would allow it to be more efficient on a given contract. Beyond direct cost-savings, a broad supply base may also result in better intangibles as the client can leverage and combine work delivered by a set of best-of-breed vendors potentially allowing the client to tap into each vendor’s innovation capabilities.
From a risk perspective, a broad vendor base also has its advantages. Having multiple suppliers may lower a client’s strategic risk by reducing lock-in. As we have argued, this is not an automatic benefit of increasing the number of vendors, as the client must also ensure that vendors are interchangeable in order to lower switching costs and associated lock-in. In our data, many sourcing managers believed that having multiple vendors with similar capabilities (e.g. several large Indian firms) involved in a given function helps each vendor become familiar with the client’s business and technological environment and hence reduces the strategic risk of being locked in. In addition, a broad vendor portfolio helps improve flexibility and reduce operational risks and one vendor’s operational failure (e.g. bankruptcy or natural disaster) will have a smaller impact on the client if switching costs are low and another vendor can take the work on. This is best seen in the cases of infrastructure work or BPO contracts with identical transactions (e.g. customer service call centres).
We have already discussed the downsides of a broad vendor portfolio, which include increased coordination and contracting costs as well as reduced economies of scale within each engagement resulting in higher production costs. It is important to note that not all functions have significant economies of scale, hence dividing work into multiple chunks may have fewer consequences. This is often the case in many types of knowledge work where economies of scale are limited to relatively small teams or groups involved in solving problems or developing new products. In any case, when the client decides to grow its vendor portfolio in order to achieve the benefits such strategy offers, it should also invest in vendor management capabilities (such as the vendor management office) to reduce ongoing coordination and contracting costs.
As we have argued, large breadth does not equal small depth. The depth is reflected in the client’s and the vendor’s mutual commitment and investment. A relationship can reach significant depth if the vendor is willing to invest time and resources in customising its knowledge and capabilities based on the client’s specific business practices and technological environment. Once the vendor acquires such strong client-specific capabilities, both the vendor’s production cost and coordination cost between the client and the vendor can be reduced. Meanwhile, intangible benefits such as innovation, operation flexibility, and quality improvement, can also be achieved. These intangible benefits are much talked about as the hallmarks of good partnership where the client gets more than what they paid for, while the vendor gets to build its capabilities by learning from and with the client.
While there are multiple benefits of close relationships with suppliers, they also have a “dark side.” This was first noted by supply chain management researchers studying the automotive industry, where close relationships developed by Japanese auto manufacturers were seen as an important part of their competitive advantage. Yet, these relationships can also result in a client being taken advantage of by vendors who feel that the relationship is so strong that the client’s switching costs are too high. Vendors would then proceed to increase prices, reduce management commitment, make poor staffing choices, etc. In response, while cultivating a deep relationship with existing vendors, the client firm should also control the cost of switching to a new vendor and consider what it would take to exit current partnerships.
Given this newly acquired understanding of how the nature of a vendor portfolio impacts outsourcing outcomes, we can see that each client can choose to use different portfolios for each function it sources. It should be noted that while there is a best-fitting multisourcing model in a certain period of time, the choice of model is not static and has to evolve. Even ABN-AMRO, which as stated took a bold step of involving five strategic vendors in IT hardware and software in 2005, has renewed its contracts in 2010 to include only three vendors. Multisourcing strategy is a process of continuous learning and adaptation. In this process, the client firm needs to keep experimenting with different configurations in search of an optimal portfolio as the business priorities of both the client and the vendors shift over time.
To conclude: differentiating between outsourcing depth and breadth and considering how each impacts drivers of outsourcing outcomes will help firms develop a strategy that results in “rightsourcing” for them.
About the Authors
Natalia Levina is Associate Professor at the Stern School of Business, New York University. She teaches courses on Global Sourcing and Open Innovation as well as Information Systems and Organizations Doctoral Seminar.
Ning Su is an Assistant Professor at the Richard Ivey School of Business at the University of Western Ontario. His research investigates global sourcing of knowledge-intensive services.