OPERATIONS RESEARCH informs ®

Vol. 57, No. 1, January–February 2009, pp. 10–18

issn 0030-364X
eissn 1526-5463
09
5701
0010 doi 10.1287/opre.1080.0628
© 2009 INFORMS

OR FORUM

The Evolution of Closed-Loop

Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

Supply Chain Research

V. Daniel R. Guide Jr.
Smeal College of Business, The Pennsylvania State University, University Park, Pennsylvania 16803,
[email protected]

Luk N. Van Wassenhove

Technology and Operations Management, INSEAD, Fontainebleau, France,
[email protected]

The purpose of this paper is to introduce the reader to the field of closed-loop supply chains with a strong business
perspective, i.e., we focus on profitable value recovery from returned products. It recounts the evolution of research in this
growing area over the past 15 years, during which it developed from a narrow, technically focused niche area to a fully
recognized subfield of supply chain management. We use five phases to paint an encompassing view of this evolutionary
process for the reader to understand past achievements and potential future operations research opportunities.
Subject classifications: closed-loop supply chains; reverse logistics; remanufacturing; value-added recovery.
Area of review: OR Forum.
History: Received March 2007; revision received January 2008; accepted May 2008.
Participate in the OR Forum discussion at http://orforum.blog.informs.org.

1. Introduction than the U.S. domestic steel industry in terms of sales

Closed-loop supply chains (CLSC) focus on taking back and employment with annual sales in excess of $53 billion
products from customers and recovering added value by (Lund 1996). Large retailers, such as Home Depot, can
reusing the entire product, and/or some of its modules, have return rates of 10% of sales, or higher, because of
components, and parts. Over the past 15 years, closed- liberal returns policies. The total value of returns can eas-
loop supply chains have gained considerable attention in ily run in the hundreds of millions of dollars for a sin-
industry and academia. This paper serves as an overview gle retailer. Stock et al. (2002) estimate the annual costs
of this evolution from a business perspective and makes of commercial returns in excess of $100 billion. Currently
some observations about future research needs. It is not very little, if any, value is recovered by the manufacturer.
intended to be a comprehensive review of published litera- Our experiences with a computer network equipment man-
ture. Rather, based on our close collaboration with compa- ufacturer showed that more than $700 million of perfectly
nies and researchers in this field, we provide our personal operational recovered products were destroyed. Hewlett-
(and perhaps biased) perspective. Packard estimates that returns cost them as much as 2% of
Today we define closed-loop supply chain management total outbound sales (Guide et al. 2006). Less than half
as the design, control, and operation of a system to maxi- of the value of those product returns is being recovered.
mize value creation over the entire life cycle of a product To make matters even worse, personal computer manufac-
with dynamic recovery of value from different types and vol- turers have short life-cycle products that can lose 1% of
umes of returns over time. This is clearly a business defini- their value per week and have high return rates (Guide et al.
tion. Fifteen years ago, the definition given by a practitioner 2006). These types of products represent a huge challenge
or researcher would have been operational and technical. for value recovery. A slow reverse supply chain that takes
This paper will describe this evolution. Our discussions are 10 weeks to put the returned product back on the mar-
restricted to value-added recovery from a purely business ket translates to a loss of (approximately) 10% of the total
perspective. However, we note that product recovery and value in that product. This far exceeds many profit mar-
reuse do serve as the foundation for the development of gins on consumer electronics, so a computer manufacturer
industrial systems that are both economically and environ- is well advised to develop competencies in fast recovery
mentally sustainable. systems.
Closed-loop supply chains have enormous economic This paper is structured as follows. We first present
potential. The remanufacturing sector is presently larger some fundamentals to understand closed-loop supply chain
10
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
Operations Research 57(1), pp. 10–18, © 2009 INFORMS 11

operations. Section 3 introduces our business perspective. By combining our two views, we can link product return
Sections 4–8 paint the evolution of research in this field types to specific recovery activities. For each type of
through a set of five phases. They describe the natural pro- product return, there is a most attractive recovery option.
gression from a technical engineering-dominated perspec- Commercial returns have barely been used and are best
tive to a holistic business model view. Please note that these reintroduced to the market as quickly as possible. The
phases should neither be taken as strictly chronological nor majority of these returns require only light repair opera-
viewed parts of a literature review. They merely serve as tions (cleaning and cosmetic). End-of-use returns may have
different lenses through which we offer the reader our per- been used intensively over a period of time and may there-
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

sonal view on the evolution of this field. Finally, we sum- fore require more extensive remanufacturing activities. The
marize the evolution of closed-loop supply chains in §9 and high variability in the use of these products may also result
make some observations about future CLSC research. in very different product disposition and remanufacturing
requirements. Ideally, one would like to acquire end-of-use
2. Closed-Loop Supply products of sufficient quality to enable profitable remanu-
Chain Fundamentals facturing. End-of-life products are predominantly techno-
logically obsolete and often worn out. This makes parts
CLSCs may be viewed with a focus on the type of returns recovery and recycling the only practical recovery alter-
or on activities. We will present both perspectives because natives (assuming that one wants to avoid landfill). Sum-
both add to an overall understanding of CLSCs. marizing, there are natural return-recovery pairs: consumer
Product returns may occur for a variety of reasons returns → repair, end-of-use returns → remanufacture, and
over the product life cycle. Commercial returns are prod- end-of-life returns → recycle. Of course, there are many
ucts returned to the reseller by consumers within 30, 60, exceptions, but our industrial experience shows that these
or 90 days after purchase (Tibben-Lembke 2004 provides a are the dominant pairings.
much finer-grained discussion of various types of commer- Before we proceed, it may be useful to reiterate that this
cial returns). End-of-use returns occur when a functional paper describes our personal view on value-added recov-
product is replaced by a technological upgrade. End-of- ery activities from a pure business perspective. This auto-
life returns are available when the product becomes tech- matically excludes a number of important subjects. Among
nically obsolete or no longer contains any utility for the others, we do not discuss green or sustainable supply chain
current user. As an illustrative example, consider the mobile issues, and we do not address end-of-life recycling and
telephone. In the United States, consumers may return a landfill avoidance (see, for example, the feature issue of the
mobile phone to the airtime provider for any reason dur- Journal of Operations Management 25(6) focused on sup-
ing a 30-day period after purchase (a commercial return). ply chain management in a sustainable environment). Like-
Furthermore, 80% of mobile phone users upgrade their wise, we do not discuss repair systems and installed base
perfectly functional mobile phones annually, making their management. Furthermore, we also exclude the literature
previous models available as an end-of-use return. Finally, on return avoidance and secondary markets for retail goods
some users of mobile phones relinquish their phone only (see Tibben-Lembke 2004 for an overview). Many authors
when it is no longer supported by the airtime provider have provided excellent contributions to all of these (and
and it becomes available as an end-of-life return (e.g., the other) topics, which one could include in a wider definition
technology is obsolete). There are also repair and warranty of closed-loop supply chain management. But our focus
returns that occur throughout, and even beyond, the product here is on value-added recovery activities, and we pro-
life cycle. It should be clear that, for consumer electron- vide our own perspective instead of engaging in a literature
ics alone, there are billions of returned products annually review. This implies by no means a value judgment about
in the United States, and therefore enormous potential for the superiority of one world view over another. It does,
value recovery. however, allow for a focused exposition of an opinion, i.e.,
Product recovery activities include used-product acqui- a forum piece.
sition, reverse logistics, product disposition (sort, test, and
grade), remanufacturing/repair, and remarketing (Guide and
3. A Business Perspective on
Van Wassenhove 2002). For example, to recover end-of-
use mobile phones, the first step is to have access to suf- Closed-Loop Supply Chains
ficient quantities of the right-quality phones at the right Taking a traditional activity-based view of the reverse sup-
price (product acquisition). The acquired mobile phones ply chain shows the key activities with a focus on indi-
must be transported to a recovery facility (reverse logistics) vidual tasks (Figure 1). The bulk of research on reverse
where they are tested, sorted, and graded (product dispo- supply chains focuses on technical and operational issues.
sition) prior to selecting the best product recovery option Obviously, remanufacturing makes no sense if techni-
(remanufacturing, repair, parts recovery, material recycling, cal/operational bottlenecks cannot be removed, so a focus
or disposal). Assuming that the optimal recovery option for on the technical activities would seem a likely place for ini-
a given mobile phone is remanufacturing, that phone then tial research attention. Even in the event that remanufactur-
needs to be sold in a secondary market (remarketing). ing is technically feasible, the potential value recovery must
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
12 Operations Research 57(1), pp. 10–18, © 2009 INFORMS

Figure 1. Activities in the reverse supply chain. Figure 2. A business process view.
Front end
Timing, quantity, quality of
used products:
Product acquisition mgmt
Return rates Front end Back end
Product Remanufacturing
returns operational
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

management issues

Reverse logistics
Remanufactured Engine
Test, sort, disposition
products market Disassemble
development Repair How can hidden
Back end
Remanufacture value be released?
Develop channels,
Remarket, secondary markets, Engine
Cannibalization could end here; however, it makes sense to now revisit
the evolution of closed-loop supply chain research in more
detail, keeping this preview in mind. To provide a mean-
exceed the costs of recovery operations. This is a necessary ingful understanding of the evolution of CLSC research,
but not sufficient condition. To make remanufacturing eco- we are going to simplify things by introducing a number
nomically attractive, one also needs adequate quantities of phases. Our discussion begins with a preview of the five
of used products of the right quality and price, at the phases.
right time, as well as a market for the recovered products. The five phases of CLSC research build on one another
In other words, one needs to go far beyond the technical in a cumulative fashion in that each phase adds new per-
and operational boundaries and take a global business pro- spectives but also deepens the study of the issues of the
cess perspective. previous research. We repeat that these phases should not
To move to a process flow perspective, we distinguish be viewed as chronological but rather as different lenses
three subprocesses (Figure 1): product returns management applied by researchers over time. Many of these phases are
(Front End), remanufacturing operational issues (Engine), concurrent. We use them mainly as a vehicle to convey our
and remanufactured products market development (Back personal perspective on the field and as a pedagogical tool
End). Because any one of these three subprocesses can be for the reader. The five phases are:
a bottleneck, the following questions that enable a process • Phase 1—The golden age of remanufacturing as a
view should be asked: technical problem.
• Does anyone want to buy remanufactured products • Phase 2—From remanufacturing to valuing the
(remanufactured products market development)? reverse-logistics process.
• Can value be recovered from returns at a reasonable • Phase 3—Coordinating the reverse supply chain.
cost (remanufacturing)? • Phase 4—Closing the loop.
• Is there sufficient access to used products (product • Phase 5—Prices and markets.
returns management)? Phase 1 clearly refers to an activity focus mainly
Our experiences suggest that very often it is not technical geared at the engine (reverse logistics and remanufactur-
constraints that matter, but rather the lack of a market for ing). Phases 2 and 3 shift the focus to a process man-
remanufactured products or the lack of used products of agement perspective and extend attention to the front and
sufficient quality at the right price and the right time. back end activities. CLSCs are typically not controlled by
To take a business perspective, we need to recognize a single actor. Our experience shows that reverse supply
that only when the three subprocesses are managed in a chains often involve many more independent players than
coordinated fashion can the value in these systems be fully forward supply chains. Indeed, one of the key issues in
realized. A lack of access to used products, or techni- designing, managing, and controlling CLSCs is the addi-
cal remanufacturing issues, or marketing and sales’ fear tional complexity that arises from this large number of
of market cannibalization can inhibit or prevent profitable actors in a decentralized system. Phase 4 moves from a
closed-loop supply chains. To make CLSCs more attractive static process management and coordination perspective to
from a business or value-creation perspective, all bottle- a dynamic one. It concentrates on dynamic system design
necks should be removed and the subprocesses smoothly over the entire product life cycle. In Phase 5, the problems
integrated (Figure 2). Only then can the hidden value be of consumer behavior and product valuation are finally rec-
released from the system. ognized. These are critical aspects previously neglected by
Research on CLSCs has evolved from examining individ- operations management researchers dominated by opera-
ual activities to considering the entire reverse supply chain tions research and industrial engineering backgrounds. We
process to finally considering closed-loop supply chains as now proceed with a more detailed exploration of each of
a potentially profitable business proposition. Our discussion the five phases.
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
Operations Research 57(1), pp. 10–18, © 2009 INFORMS 13

4. Phase 1: The Golden Age of (de Ron and Penev 1995), or design of minimum-cost recy-
Remanufacturing cling networks and the reduction of environmental impact
(Bloemhof-Ruwaard et al. 1996, Fleischmann et al. 2001).
In the early 1990s, remanufacturing was an almost com-
Please note that we do not further discuss the many excel-
pletely neglected research area, despite the size of the
lent contributions in this research stream driven by the
industry. Robert Lund’s early estimates of the size of the
environmental take-back directives of the EU (e.g., Dekker
remanufacturing industry at this time suggested that it was et al. 2004) because our focus is on value-added recovery
larger than what remained of the domestic U.S. steel indus- from a profit perspective.
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

try in terms of employment and gross sales (Lund 1996). It is important to emphasize that the early research
Despite this fact, most academics dismissed remanufactur- on CLSCs took two fundamentally different approaches:
ing as insignificant, with no relevance for the larger business market-driven (profit maximization) and waste-stream (cost
community. As the editor of a major decision sciences jour- minimization). These different world views naturally led
nal told one of us in the early 1990s, “nobody cares and to very different research interests in the United States
the issues are not mainstream research.” Fortunately, not all and Europe. However, some companies in Europe, such as
journal editors felt this way; Graham Rand, editor of the Xerox, decided to be proactive and go beyond legislation.
Journal of the Operational Research Society at the time, Xerox took a business economics perspective when it intro-
encouraged submissions addressing remanufacturing issues. duced its “green” line of remanufactured copiers. This led
Remanufacturing has existed for centuries, typically for to major changes in the production-distribution network as
high-value and low-volume items, such as locomotive well as in other functions, such as design and marketing
engines and aircraft (product life extension). The primary and sales (Thierry et al. 1995). Studying the developments
difficulty in remanufacturing large, complex items is the at Xerox, a group at Erasmus University (Rotterdam, The
scale of the problem. Products are often composed of tens Netherlands) realized that remanufacturing was not only
of thousands of components and parts. The disassembly, about minimizing the cost of compliance, but rather a more
remanufacturing, and reassembly of such products is a tech- fundamental business and value-creation issue. It was after
nical challenge with respect to shop flow control, adequate this point that a small core of researchers on both sides of
testing of critical parts (for safety reasons), and coordina- the Atlantic begin to focus on value creation.
tion of parts at the reassembly point. The earliest work The research focus during this golden age of remanufac-
by Lund (1984) at the World Bank viewed remanufac- turing was clearly on getting the engine right (reverse logis-
turing as a pathway for developing nations to get tech- tics, product disposition, and remanufacturing). It may best
nical know-how and also touted the energy savings from be described as (1) focused on cost minimization (either to
remanufactured goods (compared to production from virgin improve profitability or to reduce the cost of compliance),
materials). and (2) activity-oriented. By activity-oriented, we mean a
A small number of researchers worked on improving focus exclusively on a particular activity such as disassem-
remanufacturing shop control and coordination (Guide bly, shop floor control, or production-distribution network
1996, Guide and Srivastava 1998). This early research was design. In the United States, this literally meant extending
often sponsored by the U.S. military where remanufactur- the life of a fighter jet. In Europe, it meant compliance to
ing of expensive assets (e.g., aircraft weapons systems and legislation at minimum cost of establishing the necessary
aircraft carriers) is a critical concern. Remanufacturing job recycling networks.
shops present especially difficult situations for planning The key findings from the golden age of remanufacturing
and control because there are high uncertainties in routings were as follows: an understanding of the complicating char-
and processing times and a need for coordination among acteristics of remanufacturing and reverse logistics, and the
the different shop areas (e.g., disassembly, remanufactur- differences with traditional operations management activi-
ing processes, and reassembly) (Guide 2000). Note that this ties (Guide 2000); the characterization of common activi-
early research was driven by the need to make operations ties in reverse supply chains (Guide and Van Wassenhove
more efficient to increase profitability of the remanufactur- 2001); and the identification of different types of prod-
ing workshops. ucts returns with their specific impact on the reverse sup-
In Europe, the research focus was quite different. Re- ply chain (Thierry et al. 1995, Guide and Van Wassenhove
verse-logistics activities came about through legislation via 2001). Two things became clear during this phase. First,
European Union (EU) directives on end-of-life products, there were many new and interesting operations manage-
such as the paper recycling directive, followed by many ment problems to solve, and, second, remanufacturing was
others like the end-of-life vehicle directive and the Waste more than a niche area.
Electrical and Electronic Equipment Directive (WEEE).
These directives were essentially geared at proper recy-
5. Phase 2: From Remanufacturing to
cling and landfill avoidance. Under these circumstances,
companies naturally focused on ways to minimize the Valuing the Reverse-Logistics Process
financial impact of compliance. Therefore, researchers log- Building on Phase 1, researchers introduced two new
ically studied subjects such as design for disassembly avenues to exploring CLSC issues. In fact, these two
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
14 Operations Research 57(1), pp. 10–18, © 2009 INFORMS

approaches represent the major research themes of this research (e.g., game theory and contracting). Game the-
phase. The first uses a classic OR activity optimization ap- ory models helped to understand the strategic implications
proach (Dekker et al. 2004). It focuses on inventory control of product recovery. Contracting is of great importance
systems (Inderfurth 1997, van der Laan et al. 1999, Toktay in CLSCs because they typically have an increased number
et al. 2000, DeCroix 2006, DeCroix et al. 2005, DeCroix of actors (e.g., third-party contract providers for reverse
and Zipkin 2005, Ferrer and Whybark 2001), reverse- logistics, product disposition, remanufacturing, and remar-
logistics networks (Fleischmann et al. 2001, 2003), hybrid keting). Examining the entire process exposed huge infor-
manufacturing/remanufacturing (Aras et al. 2006), value of mation asymmetries and incentive misalignment issues in
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

information (Ferrer and Ketzenberg 2004, Ketzenberg et al. the reverse supply chain, hence the research interest in
2006), lot sizing for remanufacturing (Atasu and Cetinkaya coordination issues (for an example, see Yadav et al. 2003).
2006, Tang and Teunter 2006, Golany et al. 2001, Beltran Phase 3 is the breakthrough of the process and value view
and Krass 2002), and remanufacturing shop/line design advocated by the business economics approach and the
(Kekre et al. 2003, Ketzenberg et al. 2003, Souza et al. extension of research beyond just operational issues. It put
2002). The second, and the one that we have long advocated, a strategic lens on the front end (e.g., product acquisi-
takes a business management view. It requires connecting
tion management), as well as the back end (e.g., channel
the subprocesses (product returns management, remanufac-
design).
turing operational issues, and remanufactured products mar-
Savaskan et al. (2004) analyzed the problem of who
ket development) and exploring this area from a business
(retailer or manufacturer) should collect the returned prod-
perspective (Guide and Van Wassenhove 2001, Guide et al.
ucts under monopoly and competitive situations. If a firm
2003).
Phase 1 introduced a duality between the U.S. market- does not properly organize its access to used products, it
driven approach and the European waste-stream-driven cannot benefit from remanufacturing. Therefore, the manu-
approach. Phase 2 introduced another duality between facturer has an interest in aligning incentives for this pur-
OR-based activity optimization and a business economics pose. Debo et al. (2005) examined incentive alignment from
approach. The OR research community, primarily REVLOG the other end. Remanufacturing requires durable compo-
(an EU-sponsored research consortium consisting of six uni- nents. However, the supplier has no incentive to increase
versities), made a tremendous contribution to the definition durability if remanufacturing by the original equipment
and solution of new OR problems. These problems arose manufacturer (OEM) translates to lower sales volumes of
from the additional specifics of product return activities. As the components. In general, there is a trade-off for the OEM
an example, consider a traditional inventory problem with between the cost of investing in durability and reduced pro-
the additional option of sourcing with remanufactured com- duction costs for future generations. There is an additional
ponents (van der Laan et al. 1999). This research provided coordination problem introduced by allowing the benefits
the necessary building blocks for this new discipline. from component reuse to be earned by the OEM at the
A small group of researchers, working closely with expense of the supplier.
industry, took the business economics approach in an effort Ferguson et al. (2006) took the collection issue a step
to help resolve the larger CLSC profitability issues, which further by acknowledging that return rates can be influ-
were not well enough understood. The business economics enced, extending the system to the behavior of the reseller,
approach sought to show managers how to reuse a finan- which can be affected by the right incentives. This intro-
cially attractive option by identifying the drivers of prof- duces marketing elements to the field and illustrates the
itability (Guide and Van Wassenhove 2002)—one of the increased scope of CLSC research.
primary drivers of profitability being a product acquisition The outputs of Phase 3 provided greater understand-
management system that proactively sources used prod- ing of downstream channel design issues (Majumder and
ucts at the optimal price and quality (Guide and Van Groenevelt 2001, Savaskan et al. 2004), upstream dura-
Wassenhove 2001, Guide et al. 2003, Aras et al. 2004, bility decisions (Debo et al. 2005), the role of trade-ins
Galbreth and Blackburn 2006). (Ray et al. 2005), the interactions between new and reman-
At the end of Phase 2, we had roadmaps for study- ufactured products (Ferrer and Swaminathan 2006), and
ing a new field. OR provided a lens for understanding the reduced reseller return rates (Ferguson et al. 2006).
technically interesting issues in subfields such as inventory
Phase 3 established CLSCs as a full-fledged supply chain
control or reverse-logistics network design. The business
subfield using a business economics approach to product
economics approach allowed insights into the business pro-
returns. The research increasingly appeared in top journals,
cess challenges to release value.
and editors started to actively solicit papers (e.g., the recent
double issue of Production and Operations Management
6. Phase 3: Coordinating the Reverse 15(3&4) on CLSCs and the feature issue of Computers &
Supply Chain OR 34(2) on reverse logistics). The research community
This is where the business economics perspective linked up acknowledged that CLSCs are not just simple extensions
with other approaches in modern operations management of existing supply chain management knowledge.
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
Operations Research 57(1), pp. 10–18, © 2009 INFORMS 15

7. Phase 4: Closing the Loop There are many different return types and volumes dur-
At this phase, the research emphasis is on global system ing, and even after, the product life cycle for which one
design for profitability. The dominating view was that prod- needs to find innovative ways to recover value. As an exam-
uct returns cost money and, therefore, firms must always ple, early in the life cycle, commercial returns may be best
minimize the costs of returns (Stock et al. 2002). This used to fill warranty demands whereas, at the end of the
view stemmed from the belief that product returns were a life cycle, product returns may best be used to meet future
nuisance or, worse, trash. Conventional wisdom demanded demand for repair parts after regular production has ceased.
Closing the loop also dictates a dynamic focus on prof-
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

efficiencies because it was not realistic to spend money on

trash. Our experiences with Hewlett-Packard and Robert itability, or value creation, over the entire product life cycle,
Bosch Tools, NA suggested otherwise. We rapidly learned considering all types of product returns.
that smart firms were spending money to make money. We Summarizing, several new insights arose in Phase 4.
found that a great deal of product returns (especially com- First, a return is not just a return; there are different types of
mercial returns) are not trash, but that if slow, cost-efficient returns over the life cycle, and products have different time
processes are used, the remaining value is soon gone, leav- sensitivities. Second, minimizing the costs of returns is not
ing the firm with nothing but costly trash. always the right perspective. Third, putting insights 1 and 2
Considering time value is critical when product life cycles together is necessary to maximize value over the entire
are short. For example, the life cycle of a PC is 3 to 4 product life cycle. Phase 4 is about the correct perspective
months, and its value deteriorates at 1% per week. Clearly, at the time of system design, which determines, to a large
speed is important, and a failure to consider time in the degree, the business success of a system. This represents a
design of the closed-loop supply chain will be costly (Black- fundamentally new path for research and questions many
burn et al. 2004, Guide et al. 2006). The drivers for CLSC of the previous assumptions about centralization of facili-
system design are the volume of returns, the marginal ties, durability, and product life cycles. Cost minimization
value of time, and the quality of returned products. Time- is not the key business objective, and sometimes firms must
sensitive products, like PCs, require a responsive (decen- spend money (e.g., on product durability or responsiveness)
tralized) CLSC. Products with low time sensitivity, such as to make money.
power tools, are best served by a cost-effective (centralized)
CLSC. The key trade-off is between the opportunity cost of 8. Phase 5: Prices and Markets
value decay and economies of scale.
The research emphasis in this phase had a big-picture Phase 5 tackled the last barriers to unleashing the true
focus. Business relevance and impact are related to getting value potential in CLSCs: fear of market cannibalization
things right up-front in the design phase. Front end, engine, and returned product valuation. Market cannibalization by
and back end of the system should be integrated. These sub- remanufactured goods is “common wisdom” at many (if not
processes are not independent: the collection rate and the all) of the OEMs we have worked with (including Hewlett-
life cycle determine how durable the product should be. It is Packard and Cisco), and this makes remarketing recovered
easy to overdesign a product when the return (collection) products a tough sell. Up to this point in time, CLSC re-
rate is low and the life cycle is short. In the first case, many searchers had assumed either perfect substitution (complete
expensively designed products do not return and therefore cannibalization) or secondary markets (no cannibalization),
cannot be reused. In the second case, the products cannot and the real world rarely operates in this fashion. Granted,
be reused because they are obsolete and are no longer sold products such as single-use cameras and refillable con-
on the market. The interesting point here is that this is inde- tainers are perfect substitutes, and remanufactured mobile
pendent of the cost savings of the pure remanufacturing phones and PCs are normally sold on parallel secondary
activities. This shows that the system design depends on markets. However, the point here is that implementation of
the interrelationships among collection rate, durability, and models can be blocked by a justified fear of cannibaliza-
life cycle, and these relationships are neither intuitive nor tion because in reality most markets are a mix of the two
linear (Geyer et al. 2007). extreme cases, and research needs to acknowledge that.
True closed-loop design focus for business relevance is The back end of a closed-loop supply chain system
the key. This requires an integrated perspective on the is traditionally outside operations management boundaries.
acquisition front end (collection rates), the engine (dura- Therefore, the problem of cannibalization has only recently
bility design decisions affecting the remanufacturing oper- been addressed in OM studies (Atasu et al. 2008, Guide
ations), and remarketing back end (the life cycle and time and Li 2007). Preliminary results show that some remanu-
value of the product). This integrated design perspective has factured branded consumer products do not seem to canni-
major implications for all functions and contradicts com- balize new sales and may serve as a strategic deterrent to
mon wisdom such as the need for centralized economies low-cost competitors (Ferguson and Toktay 2006).
of scale return logistics. It also implies the recognition Up to Phase 4, researchers had also ignored product dif-
of many independent actors that need to be coordinated fusion over the life cycle. Diffusion patterns of the new
in order to unleash the potential business value. product dictate the timing and quantity of product returns,
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
16 Operations Research 57(1), pp. 10–18, © 2009 INFORMS

i.e., the life cycle of the remanufactured product. The dif- from product recovery. We used five phases to describe
fusion patterns of the new and remanufactured product are this evolution. They mainly served as a pedagogical vehi-
clearly not independent because the remanufactured prod- cle to improve understanding. Obviously, these phases did
uct sales can cannibalize new product sales, which in turn not really happen in an exact chronological order and were
influences return patterns (Debo et al. 2006). This intro- overlapped in time. While new perspectives emerged, the
duces interesting new dynamics and research questions old ones were given further in-depth attention. One can
like the optimal timing of remanufactured product intro- think of this as building a sand cone with a set of layers
duction on the market. The addition of diffusion insights, built on top of one another, which collectively have shaped
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

along with cannibalization, increases the credibility of our a new discipline. CLSCs are rapidly growing in importance
research results to the business community. We note that in industry, and this is one of the rare fields where aca-
the diffusion and cannibalization issues also introduce the demic research is ahead of industrial practice. There are
need for careful incentive alignment with the sales force. huge opportunities for increased impact with great unsolved
Rewards for selling new products are typically larger than problems ahead (Flapper et al. 2005).
for remanufactured ones, and this puts a constraint on the Summarizing our five phases, we started with provid-
perceived profitability of remanufacturing operations. ing a framework for analysis, identifying common activities
In the same vein, valuing product returns is problem- and types of returns in CLSCs and some dominant return-
atic because it obviously also determines perceived prof- recovery activity pairs. For example, end-of-use returns
itability from remanufacturing. There is no agreement in are normally best recovered via remanufacturing. Phase 1
practice on how returned products should be valued, and research focused on individual activities in the reverse sup-
the accounting literature has not addressed this problem. ply chain and established key differences with traditional
In practice, there are two extreme views. Fatalists regard OR problems in production and inventory control. For
product returns as a loss. By charging an artificially high example, in CLSCs, supply of used products is constrained
transfer price for the returned product to the remanufactur- by sales of new products, whereas supplies in traditional
ing department, the latter indeed has a hard time making models are unlimited. In Phase 2, one set of researchers
a profit. It makes product-recovery operations cost centers further developed the OR/IE stream, focusing on solving
where managers can, at best, lose less money by improv- the newly identified problems. However, some researchers
ing operations. Optimists regard product returns as a sunk shifted their attention from a local cost-minimization per-
cost (transfer price is the acquisition cost) and a potential spective to a business process view in an attempt to iden-
source of profit. In many cases, a firm can sell the products tify the drivers of profitability. Phase 3 acknowledged the
for a much higher price than the sum of the acquisition and multitude of actors in the reverse supply chain and tackled
remanufacturing costs. The two views lead to fundamen- the resulting coordination and incentive alignment issues.
tally different attitudes and behaviors because one focuses It enlarged the research perspective to include upstream
on cutting one’s losses whereas the other maximizes prof- durability decisions and downstream channel design issues.
itability. Depending on the view a firm takes, its beliefs Phase 4 took an integrative approach to the design of
about the viability of actively engaging in remanufacturing CLSCs, with a focus on profitability over the entire life
will clearly differ, as will its willingness to invest in it. cycle. It identified the value of time as an important driver
Whereas cannibalization is a real issue, in addition to being of the reverse network design and analyzed integrating all
an emotional one, in accounting terms there is no discus- product returns over the entire life cycle of the product.
sion. Fatalists are plainly wrong, and the problem is really Finally, Phase 5 extended research attention outside of the
to change incorrect accounting habits in industry. Research traditional comfort zone of OR-based operations manage-
into these issues has barely started. We refer to Toktay and ment, integrating accounting issues such as returned prod-
Wei (2005) for a preliminary discussion. uct valuation and marketing issues like cannibalization.
Phase 5 has begun to link other disciplines (i.e., mar- Throughout the phases of its evolution, CLSC research
keting and accounting) to the OM perspective. If prices moved from solving isolated OR subproblems, in what was
and markets are not fully understood, they become barri- essentially a poorly structured and unmapped field, to get-
ers, no matter how well the operational system is designed. ting a grip on the key drivers of profitability in CLSCs.
This research phase has barely begun to investigate these
Simultaneously, CLSCs have become increasingly impor-
issues. A thorough understanding clearly requires an inter-
tant in industrial practice, given diminishing life cycles,
disciplinary approach. Unfortunately, this is hampered by a
commoditization of products, and decreasing profit mar-
current lack of interest from the accounting and marketing
gins. Poorly designed reverse supply chains with no link to
research communities.
the forward business may rapidly destroy value.
Many challenging OR problems have been defined but
9. The Road Forward remain unsolved. Hence, this is a fruitful area for devel-
This paper has discussed the evolution of CLSCs from a opment of more sophisticated OR models. These mod-
technical focus on individual activities to a discipline tak- els should not necessarily be more complex, but rather
ing a holistic business process approach to releasing value more integrative, linking various disciplinary perspectives
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
Operations Research 57(1), pp. 10–18, © 2009 INFORMS 17

to provide practical solutions to the design, control, and brought together European and U.S. researchers in 1996
operation of profitable CLSCs. However, as experience has for the First International Symposium on Reuse and again
shown, given a rich area for the development of OR models in 1999 for a second meeting (de Ron and Flapper 1996,
does not mean that industry-relevant issues are the ones that 1999). The REVLOG working group, led by Erasmus
academics will pursue. In particular, we have seen a wealth University (1997–2002), brought together researchers from
of manuscripts focused on slight technical refinements to across the EU to work on reverse-logistics problems from
existing OR models or that address artificial problems (e.g., an OR perspective (Dekker et al. 2004). Finally, we need
end-of-use returns paired with remanufacturing). A focus to mention the series of workshops on Business Aspects
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

on technical extensions can trap research in a cycle where of Closed-Loop Supply Chains, of which we organized
each successive model yields a more elegant solution but the first four (2001–2005) with support from the Carnegie
with little or no connection to the larger business issues Bosch Institute and the U.S. National Science Foundation.
(i.e., profitable business models). At best, these paths of These workshops led to a number of collaborative research
research reduce our potential impact on business practice efforts, including a book (Guide and Van Wassenhove
and, at worst, could make the area irrelevant. OR models 2003a), several feature issues of journals (Interfaces 33(6),
are needed that keep the business model perspective rather California Management Review 46(2), and Production
than optimizing an isolated part of the problem. and Operations Management 15(3&4)), and many papers.
Academics should become familiar with industrial CLSC Many other authors have contributed to the development of
practice and current problems. Researchers can easily be- CLSC research from a variety of perspectives. Rogers and
come engaged with firms currently struggling with plan- Tibben-Lembke (1999) and Stock et al. (2002) clearly rec-
ning, organizing, and controlling their CLSCs. Another ognized the strategic value of reverse logistics, and Lund
pathway to understanding industry-relevant issues is to (1984) explored the technical problems in remanufacturing.
partner with another academic who has contacts and expe- We do believe that we were the first to take a business
rience with practitioners. Working with industry is difficult perspective on CLSCs and that our approach has led to rel-
and time consuming, but the potential rewards are enor- evant research and useful managerial insights. It is clear
mous. We also note that many of the previously reported that CLSC research has matured over the years and has
industry-relevant issues remain unaddressed by academic become a mainstream subarea in the supply chain field.
research (Thierry et al. 1995, Guide 2000, Guide and Van
Wassenhove 2001).
Finally, there is a strong need for interdisciplinary re- References
search with marketing (e.g., cannibalization) and account- Aras, N., T. Boyaci, V. Verter. 2004. The effect of categorizing returned
ing (e.g., valuation) to validate the assumptions that many products in remanufacturing. IIE Trans. 36 319–331.
of the early models are based on. Many assumptions, such Aras, N., V. Verter, T. Boyaci. 2006. Coordination and priority decisions
in hybrid manufacturing/remanufacturing systems. Production Oper.
as perfect substitution, are rapidly becoming institutional- Management 15 528–543.
ized, and this can reduce modeling efforts to elegant solu- Atasu, A., S. Cetinkaya. 2006. Lot sizing for optimal collection and use
tions addressing nonexistent problems. of remanufacturable returns over a finite life cycle. Production Oper.
We acknowledge that the works of many people and Management 15 473–487.
groups have not been explicitly recognized in our descrip- Atasu, A., M. Sarvary, L. N. Van Wassenhove. 2008. Remanufacturing as
tion of the evolution of CLSC research. The reason for this a marketing strategy. Management Sci. 54(10) 1731–1746.
is somewhat simple: we can best discuss what we know Beltran, J., D. Krass. 2002. Dynamic lot sizing with returning items and
disposals. IIE Trans. 34 437–448.
well (our work), and this helps us think more clearly and,
Blackburn, J., V. D. R. Guide Jr., G. Souza, L. N. Van Wassenhove. 2004.
hopefully, receive useful feedback from others. After all, Reverse supply chains for commercial returns. California Manage-
this is an opinion piece and not a literature review. Once ment Rev. 46(2) 6–23.
again, this paper does not imply a value judgment on the Bloemhof-Ruwaard, J., L. N. Van Wassenhove, H. Gabel, P. Weaver. 1996.
quality or relevance of (other) research work on CLSCs, An environmental life cycle optimization model for the European
although we do of course believe that it is important to keep pulp and paper industry. Omega 24 615–629.
a business focus in any work that claims to be relevant to Debo, L., B. Toktay, L. N. Van Wassenhove. 2005. Market segmentation
and production technology selection for remanufacturable products.
industry. Management Sci. 51 1193–1205.
The early work by Rogers and Tibben-Lembke (1999) Debo, L., B. Toktay, L. N. Van Wassenhove. 2006. Joint life-cycle dynam-
clearly identified the key role of logistics in commer- ics of new and remanufactured products. Production Oper. Manage-
cial returns and documented many practices, and Tibben- ment 15 498–513.
Lembke (2004) discusses the development of secondary DeCriox, G. 2006. Optimal policy for a multiechelon inventory system
with remanufacturing. Oper. Res. 54 523–543.
markets for retail goods. We would be remiss if we did
DeCroix, G., P. Zipkin. 2005. Inventory management for an assembly
not mention several groups that we have both been fortu-
system with product or component returns. Management Sci. 51
nate enough to be involved with. First, the early workshops 1250–1265.
organized by S. D. P. Flapper and A. de Ron at Eindhoven DeCroix, G., J.-S. Song, P. Zipkin. 2005. A series system with returns:
University of Technology (Eindhoven, The Netherlands) Stationary analysis. Oper. Res. 53 350–362.
 Guide and Van Wassenhove: The Evolution of Closed-Loop Supply Chain Research
18 Operations Research 57(1), pp. 10–18, © 2009 INFORMS

Dekker, R., M. Fleischmann, K. Inderfurth, L. N. Van Wassenhove, eds. Guide, V. D. R., Jr., L. N. Van Wassenhove, eds. 2006. Feature issue on
2004. Reverse Logistics
Quantitative Models for Closed-Loop Supply closed-loop supply chains. Production Oper. Management 15(3&4).
Chains. Springer-Verlag, Berlin. Guide, V. D. R., Jr., R. Teunter, L. N. Van Wassenhove. 2003. Match-
de Ron, A., S. D. P. Flapper, eds. 1996. Proc. First Internat. Working ing supply and demand to maximize profits from remanufacturing.
Seminar on Reuse, Eindhoven University of Technology, Eindhoven, Manufacturing Service Oper. Management 5 303–316.
The Netherlands. Guide, V. D. R., Jr., G. C. Souza, L. N. Van Wassenhove, J. D. Blackburn.
de Ron, A., S. D. P. Flapper, eds. 1999. Proc. Second Internat. Working 2006. Time value of commercial product returns. Management Sci.
Seminar on Reuse, Eindhoven University of Technology, Eindhoven, 52 1200–1214.
The Netherlands. Inderfurth, K. 1997. Simple optimal replenishment and disposal poli-
Downloaded from informs.org by [129.93.16.3] on 18 August 2014, at 10:14 . For personal use only, all rights reserved.

de Ron, A., K. Penev. 1995. Disassembly and recycling of electronic con- cies for a product recovery system with leadtimes. OR Spektrum 19
sumer products: An overview. Technovation 15(6) 363–374. 111–122.
Ferguson, M., B. Toktay. 2006. The effect of competition on recovery Jayaraman, V., R. Klassen, J. Linton, eds. 2007. Feature section on supply
strategies. Production Oper. Management 15 351–368. chains in a sustainable environment. J. Oper. Management 26(6).
Ferguson, M., V. D. R. Guide Jr., G. Souza. 2006. Supply chain coordina- Kekre, S., U. Rao, J. Swaminathan, Z. Zhang. 2003. Reconfiguring a
tion for false failure returns. Manufacturing Service Oper. Manage- remanufacturing line at Visteon, Mexico. Interfaces 33(6) 30–43.
ment 8 376–393. Ketzenberg, M., G. C. Souza, V. D. Guide Jr. 2003. Mixed assembly
Ferrer, G., M. Ketzenberg. 2004. Value of information in remanufacturing and disassembly operations for remanufacturing. Production Oper.
complex products. IIE Trans. 36 265–277. Management 12 320–335.
Ferrer, G., J. Swaminathan. 2006. Managing new and remanufactured Ketzenberg, M., E. Van der Laan, R. H. Teunter. 2006. Value of informa-
products. Management Sci. 52 15–26. tion in closed loop supply chains. Production Oper. Management 15
Ferrer, G., C. Whybark. 2001. Material planning for a remanufacturing 393–406.
facility. Production Oper. Management 10 112–124. Lund, R. T. 1984. Remanufacturing
The Experience of the United States
Flapper, S. D. P., J. van Nunen, L. N. Van Wassenhove, eds. 2005. Man- and Implications for Developing Countries. World Bank, Washington,
aging Closed-Loop Supply Chains. Springer, Berlin. D.C.
Fleischmann, M., J. van Nunen, B. Gräve. 2003. Integrating closed-loop Lund, R. T. 1996. The Remanufacturing Industry
Hidden Giant. Boston
supply chains and spare-parts management at IBM. Interfaces 33 University, Boston.
44–56. Majumder, P., H. Groenevelt. 2001. Competition in remanufacturing. Pro-
Fleischmann, M., P. Beullens, J. Bloemhof-Ruwaard, L. N. Van duction Oper. Management 10 125–141.
Wassenhove. 2001. The impact of product recovery on logistics net- Ray, S., T. Boyaci, N. Aras. 2005. Optimal prices and trade-in rebates
work design. Production Oper. Management 10 156–173. for durable, remanufacturable products. Manufacturing Service Oper.
Galbreth, M., J. D. Blackburn. 2006. Optimal acquisition and sort- Management 7 208–228.
ing policies for remanufacturing. Production Oper. Management 15 Rogers, D., R. Tibben-Lembke. 1999. Going Backwards
Reverse Logistics
384–392. Trends and Practices. RLEC Press, Pittsburgh.
Geyer, R., L. N. Van Wassenhove, A. Atasu. 2007. The impact of limited Savaskan, C., S. Bhattacharya, L. N. Van Wassenhove. 2004. Closed-loop
component durability and finite life cycles on remanufacturing profit. supply chain models with product remanufacturing. Management Sci.
Management Sci. 53 88–100. 50 239–252.
Golany, B., J. Yang, G. Yu. 2001. Economic lot-sizing with remanufac- Souza, G., M. Ketzenberg, V. D. R. Guide Jr. 2002. Capacitated reman-
turing options. IIE Trans. 33 995–1004. ufacturing with service level constraints. Production Oper. Manage-
Guide, V. D. R., Jr. 1996. Scheduling using drum-buffer-rope in a reman- ment 11 231–248.
ufacturing environment. Internat. J. Production Res. 34 1081–1091. Stock, J., T. Speh, H. Shear. 2002. Many happy (product) returns. Harvard
Guide, V. D. R., Jr. 2000. Production planning and control for remanu- Bus. Rev. 80(7) 16–17.
facturing: Industry practice and research needs. J. Oper. Management Tang, O., R. H. Teunter. 2006. Economic lot scheduling problem with
18 467–483. returns. Production Oper. Management 15 488–497.
Guide, V. D. R., Jr., K. Li. 2007. The potential for cannibalization of Thierry, M., M. Salomon, J. A. E. E. van Nunen, L. N. Van Wassenhove.
new product sales by remanufactured products. Working paper, Smeal 1995. Strategic issues in product recovery management. California
College of Business, The Pennsylvania State University, University Management Rev. 37 114–135.
Park. Tibben-Lembke, R. 2004. Strategic use of the secondary market for retail
Guide, V. D. R., Jr., R. Srivastava. 1998. Inventory buffers in recoverable consumer goods. California Management Rev. 46 90–104.
manufacturing. J. Oper. Management 16 551–568. Toktay, B., D. Wei. 2005. Cost allocation in manufacturing-
Guide, V. D. R., Jr., L. N. Van Wassenhove. 2001. Managing prod- remanufacturing operations. Working paper, College of Management,
uct returns for remanufacturing. Production Oper. Management 10 Georgia Institute of Technology, Atlanta.
142–155. Toktay, B., L. Wein, S. Zenios. 2000. Inventory management of remanu-
Guide, V. D. R., Jr., L. N. Van Wassenhove. 2002. The reverse supply facturable products. Management Sci. 46 1412–1426.
chain. Harvard Bus. Rev. 80(2) 25–26. van der Laan, E., M. Salomon, R. Dekker, L. N. Van Wassenhove. 1999.
Guide, V. D. R., Jr., L. N. Van Wassenhove, eds. 2003a. Business Aspects Inventory control in hybrid systems with remanufacturing. Manage-
of Closed-Loop Supply Chains. Carnegie Mellon University Press, ment Sci. 45 733–747.
Pittsburgh. Verter, V., T. Boyaci, eds. 2007. Feature issue on reverse logistics. Comput.
Guide, V. D. R., Jr., L. N. Van Wassenhove, eds. 2003b. Special section & OR 34(2).
on closed-loop supply chains. Interfaces 33(6) 1–2. Yadav, P., D. Miller, C. Schmidt, R. Drake. 2003. McGriff Treading Com-
Guide, V. D. R., Jr., L. N. Van Wassenhove, eds. 2004. Feature issue on pany implements service contracts under shared savings. Interfaces
closed-loop supply chains. California Management Rev. 46(2). 33 18–29.