A framework for supply chain performance measurement

ARTICLE IN PRESSInt. J. Production Economics 87 (2004) 333–347A framework for supply chain performance measurementA. Gunasekaran a, *, C. Patel b , Ronald E. McGaughey ca Department of Business Administration, University of Illinois at Springfield, One University Plaza, Springfield, IL 62703-5407, USAb Ace InfoTech, Inc., 406 Wellington Drive, Streamwood, IL 60107, USAc Department of Management Information Systems, The University of Central Arkansas, Conway, AR 72035-0001, USAAbstractSupply chain management (SCM) has been a major component of competitive strategy to enhance organizationalproductivity and profitability. The literature on SCM that deals with strategies and technologies for effectivelymanaging a supply chain is quite vast. In recent years, organizational performance measurement and metrics havereceived much attention from researchers and practitioners. The role of these measures and metrics in the success of anorganization cannot be overstated because they affect strategic, tactical and operational planning and control.Performance measurement and metrics have an important role to play in setting objectives, evaluating performance,and determining future courses of actions. Performance measurement and metrics pertaining to SCM have not receivedadequate attention from researchers or practitioners. We developed a framework to promote a better understanding ofthe importance of SCM performance measurement and metrics. Using the current literature and the results of anempirical study of selected British companies, we developed the framework presented herein, in hopes that it wouldstimulate more interest in this important area.r 2003 Elsevier B.V. All rights reserved.Keywords: Supply chain; Performance measurements; Metrics; Empirical analysis; Framework1. IntroductionBy the late 1980s, outsourcing in US industriescontributed to nearly 60% of the total productcost (Ballou, 1992). In the UK, a survey showedthat 40% of the UK’s gross domestic product wasspent on distribution and logistics related activities(Department of Trade and Industry, UK, 1990).Such findings and developments present significantvisible impact of distribution, purchasing, andsupply management on company assets. Managers*Corresponding author. Tel.: +1-217-206-7927; fax: +1-217-206-7543.E-mail address: Gunasekaran.angappa@uis.edu(A. Gunasekaran).in many industries, especially those in manufacturing,are trying to better manage supply chains.Important techniques/methodologies like just-intime(JIT), total quality management, lean production,computer generated enterprise resourceplanning schedule (ERP) and Kaizen have beenembraced. The concept of supply chain management(SCM), according to Thomas and Griffin(1996) represents the most advanced state in theevolutionary development of purchasing, procurementand other supply chain activities. At theoperational level, this brings together functionsthat are as old as commerce itself—seeking goods,buying them, storing them and distributing them.At the strategic level, SCM is a relatively new andrapidly expanding discipline that is transforming0925-5273/$ - see front matter r 2003 Elsevier B.V. All rights reserved.doi:10.1016/j.ijpe.2003.08.003

ARTICLE IN PRESS334A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347the way that manufacturing and non-manufacturingoperations meet the needs of their customers.Development of cross-functional teams alignsorganisations with process oriented structure,which is much needed to realise a smooth flow ofresources in a supply chain. As suggested by Trentand Monczka (1994), such teams promote improvedsupply chain effectiveness. They minimiseor eliminate functional and departmental boundariesand overcome the drawbacks of specialisation,which according to Fawcett (1995), candistribute the knowledge of all value addingactivities such that no one, including upper levelmanagers, has complete control over the process.Such teams helped in the formation of modernsupply chains by promoting greater integrationof organisations with their suppliers andcustomers.Supplier partnerships and strategic alliancesrefer to the co-operative and more exclusiverelationships between organisations and their upstreamsuppliers and downstream customers. Todaymany firms have taken bold steps to breakdown both inter and intra firm barriers to formalliances, with the objective of reducing uncertaintyand enhancing controlof supply anddistribution channels. Such alliances are usuallycreated to increase the financialand operationalperformance of each channelmember throughreductions in totalcost and inventories andincreased sharing of information (Maloni andBenton, 1997). Rather than concerning themselvesonly with price, manufacturers are looking tosuppliers to work co-operatively in providingimproved service, technological innovation andproduct design. This development has produced asignificant impact by expanding the scope of SCMthrough greater integration of suppliers withorganisations.The growth and development of SCM is notdriven only by internal motives, but by a numberof externalfactors such as increasing globalisation,reduced barriers to internationaltrade, improvementsin information availability, and environmentalconcerns.Furthermore, computer generatedproduction schedules, increasing importanceof controlling inventory, government regulationsand actions such as the creation of a singleEuropean market, and the guidelines of GATTand WTO have provided the stimulus for developmentof and existing trends in SCM. Supplychain integration is needed to manage and controlthe flow in operating systems. Such flow controlisassociated with inventory controland activitysystem scheduling across the whole range ofresource and time constraints. Supplementing thisflow control, an operating system must try to meetthe broad competitive and strategic objectives ofquality, speed, dependability, flexibility and cost(Slack et al., 1995; Gunasekaran et al., 2001; DeToni and Tonchia, 2001). Controlis also essentialas both customer needs and supply chain performancemight change with time.To meet objectives, the output of the processesenabled by the supply chain must be measured andcompared with a set of standards. In order to becontrolled, the process parameter values need to bekept within a set limit and remain relativelyconstant. This will allow comparison of plannedand actualparameter values, and once done, theparameter values can be influenced throughcertain reactive measures in order to improve theperformance or re-align the monitored value tothe defined value. For example, an analysis of thelayout of facilities could reveal the cause of longdistribution time, high transportation and movementcosts and inventory accumulation. Usingsuitable approaches like re-engineering facilities,problems can be tackled and close monitoring andsubsequent improvements can be possible fromanalysis of the new design. Thus, control ofprocesses in a supply chain is crucial in improvingperformance and can be achieved, at least in part,through measurement. Well-defined and controlledprocesses are essential to better SCM.There are number of conceptualframeworksand discussions on supply chain performancemeasurements in the literature; however, there isa lack of empirical analysis and case studies onperformance metrics and measurements in asupply chain environment. We will discuss thebackground for the research, review the selectedliterature on supply chain performance metricsand measurements, develop a framework based onthe literature and an empirical analysis, andfinally, summarize the findings and conclusions.

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 3352. Background for researchIn this section, the literature is used in describingthe generalcontext within which measurementof supply chain performance is undertaken. Theworks of various authors are used in establishingthe need for supply chain performance measurementand to describe in generalterms how it should beaddressed—emphasis is on measurement systemsand approaches as opposed to specific measures.The strategic, operational and tactical levels arethe hierarchies in function, wherein policies andtrade-offs can be distinguished and suitable controlexerted(Ballou, 1992). According to Rushtonand Oxley (1989), such a hierarchy is based on thetime horizon for activities and the pertinence ofdecisions to and influence of different levels ofmanagement. The strategic level measures influencethe top level management decisions, veryoften reflecting investigation of broad basedpolicies, corporate financial plans, competitivenessand level of adherence to organisational goals. Thetactical level deals with resource allocation andmeasuring performance against targets to be metin order to achieve results specified at the strategiclevel. Measurement of performance at this levelprovides valuable feedback on mid-level managementdecisions. Operational level measurementsand metrics require accurate data and assess theresults of decisions of low level managers. Supervisorsand workers are to set operationalobjectivesthat, if met, will lead to the achievement oftacticalobjectives.Many firms look to continuous improvement asa toolto enhance their core competitiveness usingSCM. Many companies have not succeeded inmaximizing their supply chain’s potential becausethey have often failed to develop the performancemeasures and metrics needed to fully integratetheir supply chain to maximize effectiveness andefficiency. Lee and Billington (1992) observed thatthe discrete sites in a supply chain do not maximizeefficiency if each pursues goals independently.They point to incomplete performance measuresexisting among industries for assessment of theentire supply chain. Measurements should beunderstandable by all supply chain members andshould offer minimum opportunity for manipulation(Schroeder et al., 1986). Performance studiesand models should be created so that organisationalgoals and achievement of those goals can bemeasured, thus allowing the effectiveness of thestrategy or techniques employed to be accessed.Most companies realise the importance offinancialand non-financialperformance measures,however they have failed to represent them in abalanced framework. According to Kaplan andNorton (1992), while some companies and researchershave concentrated on financialperformancemeasures, others have concentrated onoperationalmeasures. Such an inequality doesnot lead to metrics that can present a clear pictureof organisationalperformance. For a balancedapproach, Maskell (1991) suggests that companiesshould understand that, while financial performancemeasurements are important for strategicdecisions and externalreporting, day to daycontrolof manufacturing and distribution operationsis often handled better with non-financialmeasures. Another area where inequality persists isdeciding upon the number of metrics to be used.Quite often companies have a large number ofperformance measures to which they continue toadd based on suggestions from employees andconsultants. They fail to realise that performanceassessment can be better addressed using a trivialfew—they are not really trivial, but instead arethose few areas most criticalto success.The metrics that are used in performancemeasurement and improvement should be thosethat truly capture the essence of organizationalperformance. A measurement system should facilitatethe assignment of metrics to where theywould be most appropriate. For effective performancemeasurement and improvement, measurementgoals must represent organisational goalsand metrics selected should reflect a balancebetween financialand non-financialmeasures thatcan be related to strategic, tactical and operationallevels of decision making and control.3. Performance measurements and metrics in SCMIn this section, the literature on performancemeasurements and metrics in SCM is reviewed.

ARTICLE IN PRESS336A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347The metrics and measures are discussed in thecontext of the following supply chain activities/processes: (1) plan, (2) source, (3) make/assemble,and (4) delivery/customer (Stewart, 1995;Gunasekaran et al., 2001).3.1. Metrics for order planning3.1.1. The order entry methodThis method determines the way and extent towhich customer specifications are converted intoinformation exchanged along the supply chain.3.1.2. Order lead-timeThe total order cycle time, called order todelivery cycle time, refers to the time elapsed inbetween the receipt of customer order untilthedelivery of finished goods to the customer. Thereduction in order cycle time leads to reduction insupply chain response time, and as such is animportant performance measure and source ofcompetitive advantage (Christopher, 1992)—itdirectly interacts with customer service in determiningcompetitiveness.3.1.3. The customer order pathThe path that an order traverses is anotherimportant measure whereby the time spent indifferent channels can be determined. By analyzingthe customer order path, non-value adding activitiescan be identified so that suitable steps can betaken to eliminate them.3.2. Evaluation of supply linkTraditionally supplier performance measureswere based on price variation, rejects on receiptand on time delivery. For many years, the selectionof suppliers and product choice were mainly basedon price competition with less attention affordedto other criteria like quality, reliability, etc. Morerecently, the whole approach to evaluating suppliershas undergone drastic change.Evaluation of suppliers: The evaluation ofsuppliers in the context of the supply chain(efficiency, flow, integration, responsiveness andcustomer satisfaction) involves measures importantat the strategic, operationaland tacticallevel.Strategic level measures include lead timeagainst industry norm, Quality level, Costsaving initiatives, and supplier pricing againstmarket.Tactical level measures include the efficiency ofpurchase order cycle time, booking in procedures,cash flow, quality assurance methodology andcapacity flexibility.Operational level measures include ability in dayto day technicalrepresentation, adherence todeveloped schedule, ability to avoid complaintsand achievement of defect free deliveries.Purchasing and supply management must analyzeon a periodic basis their supplier abilities tomeet the firm’s long-term needs. The areas thatneed particular attention include the supplier’sgeneral growth plans, future design capability inrelevant areas, role of purchasing and supplymanagement in the supplier’s strategic planning,potentialfor future production capacity andfinancialability to support such growth (Fisher,1997). Supply chain partnership is a collaborativerelationship between a buyer and seller whichrecognises some degree of interdependence and cooperationon a specific project or for a specificpurchase agreement (Ellram, 1991; van Hoek,2001). Such a partnership emphasises direct,long-term association, encouraging mutual planningand problem solving efforts (Maloni andBenton, 1997). Supplier partnerships have attractedthe attention of practitioners and researchers(Macbeth and Ferguson, 1994; Ellram, 1991;Graham et al., 1994). All have contended thatpartnership formation is vitalin supply chainoperations and as such for efficient and effectivesourcing. Partnership maintenance is no lessimportant. Performance evaluation of buyers orsuppliers is simply not enough—relationships mustbe evaluated.The parameters that need to be considered in theevaluation of partnerships are the ones thatpromote and strengthen them. For example, thelevel of assistance in mutual problem solving isindicative of the strength of supplier partnerships.Partnership evaluation based on such criteria willresult in win–win partnerships leading to moreefficient and more thoroughly integrated supplychains.

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 3373.3. Measures and metrics at production levelAfter the order is planned and goods sourced,the next step in to make/assemble products. This isthe activity carried out by organisations that ownproduction sites, and their performance has amajor impact on product cost, quality, speed ofdelivery and delivery reliability, and flexibility(Mapes et al., 1997; Slack et al., 1995). As it isquite an important part of the supply chain,production needs to be measured and continuouslyimproved. Suitable metrics for the productionlevel are as follows:Range of product and services: According to Mapeset al. (1997), a plant that manufactures a broadproduct range is likely to introduce new productsmore slowly than plants with a narrow productrange. Plants that can manufacture a wide range ofproducts are likely to perform less well in the areasof value added per employee, speed and deliveryreliability. This clearly suggests that product rangeaffects supply chain performance.Capacity utilization: From the above assertion, it isclear that the role-played by capacity in determiningthe level of activities in a supply chain is quiteimportant. According to Slack et al. (1995), of themany aspects of production performance, capacityutilization directly affects the speed of response tocustomer demand through its impact on flexibility,leadtime and deliverability.Effectiveness of scheduling techniques: Schedulingrefers to the time or date on or by which activitiesare to be undertaken. Such fixing determines themanner in which resources will flow in anoperating system, the effectiveness of which hasan important impact on production and thussupply chain performance. For example, schedulingtechniques such as JIT, MRP and ERP haveimplications on purchasing, throughput time andbatch size. In case of the supply chain, sincescheduling depends heavily on customer demandsand supplier performance, the scheduling toolsshould be viewed in that context (Little et al.,1995).3.4. Evaluation of delivery linkThe link in a supply chain that directly impactscustomers is delivery. It is a primary determinantof customer satisfaction; hence, measuringand improving delivery is always desirable toincrease competitiveness. Delivery by its verynature takes place in a dynamic and ever-changingenvironment, making the study and subsequentimprovement of a distribution system difficult. Itshould be noted that it is not an easy matter toanticipate how changes to one of the majorelements within a distribution structure willaffect the system as a whole (Rushton and Oxley,1989).3.4.1. Measures for delivery performanceevaluationAccording to Stewart (1995), an increase indelivery performance is possible through a reductionin leadtime attributes. Another importantaspect of delivery performance is on-time delivery.On-time delivery reflects whether perfect deliveryhas taken place or otherwise and is also a measureof customer service level. A similar concept, ontime order fill, was used by Christopher (1992),describing it as a combination of delivery reliabilityand order completeness. Another aspect ofdelivery is the percentage of finished goods intransit, which if high signifies low inventoryturns, leading to unnecessary increases in tied upcapital. Various factors that can influence deliveryspeed include vehicle speed, driver reliability,frequency of delivery, and location ofdepots. An increase in efficiency in these areas canlead to a decrease in the inventory levels (Novich,1990).Number of faultless notes invoiced: An invoiceshows the delivery date, time and conditionunder which goods were received. By comparingthese with the previously made agreement, itcan be determined whether perfect delivery hastaken place or not, and areas of discrepancycan be identified so that improvements can bemade.Flexibility of delivery systems to meet particularcustomer needs: This refers to flexibility in meetinga particular customer delivery requirement at an

ARTICLE IN PRESS338A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347agreed place, agreed mode of delivery and withagreed upon customised packaging. This type offlexibility can influence the decision of customersto place orders, and thus can be regarded asimportant in enchanting and retaining customers(Novich, 1990).3.4.2. Total distribution costPerhaps the most important research concerninglogistics is going on in the area of design ofefficient and cost effective distribution systems.For this, an understanding of totaldistributioncost is essential, so that proper trade-offs can beapplied as a basis for planning and reassessment ofdistribution systems. The urgency of dealing withtransportation cost was highlighted by Thomasand Griffin (1996), who argued that since transportationcost accounts for more than half of thetotallogistics cost, more active research is neededin the area. To dealwith distribution costs,measuring individualcost elements together withtheir impact on customer service encourages tradeoffsthat lead to a more effective and efficientdistribution system.3.5. Measuring customer service and satisfactionTo a world class organisation, a happy andsatisfied customer is of the utmost importance. Ina modern supply chain customers can reside nextdoor or across the globe, and in either case theymust be well served. Without a contented customer,the supply chain strategy cannot be deemedeffective. Lee and Billington (1992) and van Hoeket al. (2001) emphasised that to assess supply chainperformance, supply chain metrics must centre oncustomer satisfaction.3.5.1. FlexibilityOf the factors by which supply chains compete,flexibility can be rightly regarded as a critical one.Being flexible means having the capability toprovide products/services that meet the individualdemands of customers. Some flexibility measuresinclude: (i) product development cycle time, (ii)machine/toolset up time, (iii) economies of scope(Christopher, 1992)—refers to the production ofsmall quantities of wider range (e.g. JIT lot size)—and (iv) number of Inventory turns.3.5.2. Customer query timeCustomer query time relates to the time ittakes for a firm to respond to a customer querywith the required information. It is not unusualfor a customer to enquire about the statusof order, potential problems on stock availability,or delivery. A fast and accurate responseto those requests is essentialin keeping customerssatisfied.3.5.3. Post transaction measures of customerserviceThe function of a supply chain does not endwhen goods are provided to the customer. Posttransaction activities play an important role incustomer service and provide valuable feedbackthat can be used to further improve supply chainperformance.3.6. Supply chain and logistics costThe efficiency of a supply chain can beassessed using the totallogistics cost—a financialmeasure. It is necessary to assess the financialimpact of broad level strategies and practicesthat contribute to the flow of products in asupply chain. Since logistics cut across functionalboundaries,care must be taken to assessthe impact of actions to influence costs in onearea in terms of their impact on costs associatedwith other areas (Cavinato, 1992). Forexample, a change in capacity has a major effecton cost associated with inventory and orderprocessing.3.6.1. Cost associated with assets and return oninvestmentSupply chain assets include accounts receivable,plant, property and equipment, and inventories.With increasing inflation and decreased liquidity,pressure is on firms to improve the productivity ofcapital—to make the assets sweat. In this regard itis essentialto determine how the cost associatedwith each asset, combined with its turnover, affectstotalcash flow time. One way to address this is by

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 339expressing it as an average days required toturn cash invested in assets employed intocash collected from a customer (Stewart, 1995).Thus, totalcash flow time can be regardedas a metric to determine the productivity ofassets in a supply chain. Once the total cashflow time is determined, this can be readilycombined with profit to provide insight intothe rate of return on investment (ROI). Thisdetermines the performance by top managementis terms of earnings on the totalcapitalinvestedin a business.With customer service requirements constantlyincreasing, effective management of inventory inthe supply chain is crucial (Slack et al., 1995). In asupply chain, the total cost associated withinventory can be broken down into the following(Stewart, 1995; Christopher, 1992; Slack et al.,1995; Lee and Billington, 1992; Levy, 1997):Opportunity cost, consisting of warehousing,capitaland storage; Cost associated with inventoryat the incoming stock level and work in progress;Service costs, consisting of cost associated withstock management and insurance; Cost of finishedgoods including those in transit; Risk costs,consisting of cost associated with pilferage,deterioration, and damage; Cost associated withscrap and rework; and Cost associated with toolittle inventory accounting for lost sales/lostproduction.3.6.2. Information processing costThis includes costs such as those associated withorder entry, order follow/updating, discounts, andinvoicing. On the basis of survey results fromvarious industries, Stewart (1995) identified informationprocessing cost as the largest contributorto total logistics cost. The role of informationtechnology is shifting from a general passivemanagement enabler through databases, to ahighly advanced process controller that canmonitor activities and decide upon an appropriateroute for information. Modern information technology,through its power to provide timely,accurate, and reliable information, has led to agreater integration of modern supply chains thanpossible by any other means (Naim, 1997;Benjamin and Wigand, 1995).4. The research methodologyThe framework presented by Gunasekaran et al.(2001) was used in developing a survey used tostudy performance measures and metrics used in asupply chain environment. A seven-page questionnaire1 was developed for collecting data. Thequestionnaire was divided into four basic sections.They are as follows: plan (including strategy),source/supply (order), produce (make/assemble),and delivery (to customer). These four categoriescorrespond to the four basic activities or processesin a supply chain—plan–source–make/assemble–delivery. The questionnaires were mailed with acover letter and addressed to the CEO of eachfirm. Targeted recipients were instructed to completethe survey themselves or refer it to anappropriate person for the same. Participants wereidentified using the ‘Kompass Register’ for UKindustries (Volumes I and II) published by theReed Business Information Ltd., West Sussex,UK. A total of 150 large companies were selectedfrom a wide range of industry settings.5. Empirical analysisOf the 150 questionnaires mailed, 21 werecompleted and returned. A breakdown of thesurvey response is shown in Fig. 1. Nearly all theresponses were received within 4 weeks of mailing.Twelve companies said that because of the largernumber of such enquiries they were unable toreply. Ten companies returned the questionnairestating that they were not suitable candidates forthe survey because of changes in their operations.The response rate was only 14%, but we felt that itwas adequate to assist us in developing ourframework.5.1. Planning performance evaluation metricsThis section deals with financial and nonfinancialstrategiclevelperformance measures.The importance of these parameters was establishedby calculating the mean of all responses and1 Available upon request from authors

ARTICLE IN PRESS340A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347Responded withcompleted questionnaire14%Said were notable to replyTable 1Ratings strategic planning metrics8%AssessmentStrategic performancemetricsPercentageimportance71%Did not respond tothe questionnaireFig. 1. Breakdown of response for the survey.7%Returned thequestionnairewithoutcompleting itranking them accordingly. The ranks were convertedto relative percentages by dividing eachrank, by the totalof a lranks for the group ofmeasures/metrics. This approach is similar to themethod used in Pareto analysis wherein problemfrequencies are converted to percentages to showrelative importance. The percentages better highlightdifferences in the importance of performancemeasures in each group (we used this approach inanalysing performance measures in all groupsdiscussed herein). We further categorized themeasures based on importance (highly important,moderately important and less important). Themethodology employed for such was similar to themethodology used in ABC inventory (inventoryitem’s annualcost is stated as a percent of totalinventory costs) to prioritize inventory managementdecisions (item cost percentages sorted indescending order and grouped into A—mostimportant, B—moderate importance, and C—lessimportant based on their contribution to totalcosts). We used this approach in analysingperformance measures in all groups discussedherein. Please note that categorizing a measure asless important does not mean it is unimportant,but rather it seems less important compared toothers in the measurement group. We believe asimilar approach could be used by managers insetting priorities in the development of a measurementsystem for supply chain performance. Oursmall sample size precluded the use of morepowerfulstatisticaltechniques. We believe ourapproach is adequate for our use of the data inframework development. A more rigorous studyHighly importantModeratelyimportantLess importantLevel of customerperceived value ofproduct16.42Variances against 14.23budgetOrder lead time 13.50Information processing 12.68costNet profit Vs12.46productivity ratioTotalcycle time 11.80Totalcash flow time 10.27Levelof energyutilisation8.64to validate the framework should employ a bettersample and more rigorous statistical techniques.The first set of measures (five non-financialandthree financial) pertain to planning, but morespecifically to strategic planning. Table 1 showsthe measures and their relative importance asdetermined by our analysis of the survey data.The importance rating survey results show thatthe level of customer perceived value of product isof the utmost importance. It was deemed highlyimportant which clearly reflects the perception ofpractitioners that customer satisfaction is paramountin importance in increasing competitiveness.The measures considered moderatelyimportant in descending order include variancesagainst budget, order lead-time, informationprocessingcost, net profit vs. productivity ratio,totalcycle time and totalcash flow time. Variancesagainst budget, information-processing cost andnet profit vs. productivity are of course financialmeasures and reflect the importance of financialmeasures in strategic planning and control—financialstability is essentialto organizationalsuccess. The other three moderately importantmeasures were order lead time, total cycle time andtotalcash flow time. Their rating further highlightsthe importance of non-financialmeasures instrategic planning and control and to subsequent

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 341Table 2Importance of order planning metricsAssessment Metrics PercentageimportanceHighly important Customer query time 19.11ModeratelyimportantProduct developmentcycle time17.37Less important Accuracy of forecasting 16.59Planning process cycle 15.90timeOrder entry methods 15.51Human resourceproductivity15.51organizational success. The only strategic planningmeasure deemed less important was level of energyutilisation which may suggest that it is not ofstrategic significance. That, of course, could varyfrom firm to firm, depending on energy cost as apercent of totalmanufacturing cost and on energyprice levels relative to the prices of other manufacturinginputs.The percentage importance (relative importance)of the strategic performance metrics clearlysuggests that non-financialmeasures of performanceare considered by practitioners to beimportant in assessing the competitiveness of anorganization. This is not to say that financialmeasures are no longer important, but rather thatnon-financialmeasures are important and necessaryin assessing a firm’s ability to compete.In Table 2, the order of priority for the orderplanning level metrics is presented. At the orderplanning level, customer query time was highlyimportant, which would seem to emphasize theimportance of customer service. Product developmentcycle time and forecasting were moderatelyimportant. These two factors relate to meetingcustomer needs and doing so in a timely fashion.Although there is no statistical evidence containedherein to prove such a link, common sensesuggests a link between these and the perceivedcustomer value of the product, rated number oneamong the strategy performance measures. Theimportance ratings of product development cycletime and forecasting measures suggests that theywarrant monitoring by management and improvementeffort. Cross-functionalteams, rapid prototyping,and concurrent engineering involvingsuppliers would seem appropriate in efforts toimprove product development cycle time. Manyalternative techniques are available for forecasting.If forecasting accuracy is a concern, firms mightexamine the techniques employed with an eyetoward improvement. Because the forecasts of allsupply chain links can influence supply chainperformance, a concerted effort by all should bemade to assure accurate forecasts. This is emphasizedby a survey participant (a machine toolmanufacturer) who said that supply chain partnersshould ‘‘Use better forecasting techniques to removeuncertainties in supply chain.’’ Many understandthe consequences of weak forecastingperformance and recognize the need to measureand improve it.By benchmarking their forecasting methodswith those of the best, a better understanding thetechniques might be gained and greater accuracyachieved. Also, by integrating production scheduleswith others in the supply chain, moreaccurate day to day demand forecast might bepossible for all links in the supply chain. Planningprocess cycle time, order entry methods, andhuman resource productivity were the less importantorder planning measures. Planning processcycle time and order entry methods could beimproved through reengineering efforts that includemultiple links in the supply chain, becausethe actions of multiple participants interact toinfluence performance in these areas. Improvementsin customer query time, product developmentcycle time and planning process cycle timemight be brought about by greater humanresource productivity, so although it was ratedlast in importance, human resource productivityshould not be dismissed as unimportant. Improvementin order entry methods, customer query time,forecasting accuracy and customer query timemight be brought about through the application ofinformation technology to increase accuracy andexpedite the flow of information throughout thesupply chain. Process cycle time can be tackled byusing techniques like single minute exchange of dieand group technology, whereby similar facilities

ARTICLE IN PRESS342A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347for production will be grouped to reduce manufacturinglead-time.5.2. Supply link evaluation metricsDue to the growing importance outsourcing,whereby firms outsource a major part of theirproducts, evaluation of supply link performance isvery important in managing the supply chain forpeak efficiency and effectiveness. In this section,the importance of performance measures/metricsin a supply chain link (includes purchasing andsupplier management activities) are rated inimportance. Based on the literature, six keyperformance indicators (KPI) pertaining to thesupplier link were included in the survey andranked by participants. These measures include:supplier delivery performance, lead-time againstindustry norm, supplier pricing against market,efficiency of purchase order cycle time, efficiencyof cashflow method, and supply booking procedures.The main objective here is to identify theKPI in supply link performance evaluation. TheKPI can be defined as the performance indicatorsthat have significant impact on the overallperformance of an organization in the areas ofstrategic, tacticaland operationalplanning andcontrol. The percentage importance ratings of thesix measures are included in Table 3.As can be seen in Table 3, supplier deliveryperformance emerged as the most importantmeasure pertaining to the evaluation of supplierperformance. It was the only highly importantmeasure. One can see from the table that it isclearly set apart from the others by its percentageimportance rating. The moderately importantmeasures in descending order are supplier leadtimeagainst industry norm, supplier pricingagainst market and efficiency of purchase ordercycle time. The less important supplier measureswere efficiency of cash flow method and supplierbooking in procedures. Most notable about thesupplier metrics is that firms regard the supplier’scapability to reliably deliver goods in a timelyfashion as more important than price. Price hasincreasing become an order qualifier rather thanan order winner. Other aspects of supplierperformance such as adherence to agreed uponTable 3Importance of supplier metricsAssessment Metrics PercentageimportanceHighly importantModeratelyimportantLess importantSupplier deliveryperformanceSupplier lead-timeagainst industry normSupplier pricing againstmarketEfficiency of purchaseorder cycle timeEfficiency of cash-flowmethodSupplier booking inprocedures23.2019.6918.3015.4212.3811.01schedules and terms of the order as well as promptdelivery of goods have become order winners.Firms would do well to not just use suppliermetrics for selection of suppliers, but rather theyshould work closely with suppliers to see that theyhave in place within their organizations, measurementsystems that will foster significant improvementin all of these areas. Such improvementcontributes to the overall success of a supplychain.5.3. Production performance evaluation metricsIn this section, supply chain production linkmetrics/measures are rated in importance. Theliterature provided the production link measures,and as with other metrics evaluated in this paper,the survey responses provided the basis for ratingthe importance of these measures. The performancemeasures for the production link includedpercentage of defects (a measure of productquality), cost per operation-hour, capacity utilization,range of product and services, and utilizationof economic order quantity. Table 4 contains themeasures and their percentage importance ratings.From the table one can see that the percentage ofdefects emerged to be the most important(24.27%), but two others, cost per operation hourand capacity utilization, were also highly important.The latter two are essentially measures of the

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 343Table 4Importance of production metricsAssessment Metrics PercentageimportanceHighly important Percentage of defects 24.27Cost per operation hour 22.51Capacity utilization 21.61ModeratelyimportantLess importantRange of products andservicesUtilization of economicorder quantity18.0113.60efficiency with which resources are used inmanufacturing (produce/assemble), and good performancein these two areas translates into lowercost per unit to manufacture products/provideservices. Efficiency of operations is important forall supply chain partners, if the elusive goal ofsupply chain optimization is to be achieved. Notethat the percentage importance of each of thesethree clearly sets them apart from the moderatelyimportant and less important measures. We shouldcaution that maximum efficiency of each partnerin all areas might not be a desirable becausetradeoffs are necessary in order to achieve a globaloptimum for the supply chain—local optimums inall parts do not necessarily lead to globaloptimization for a system.The only measure rated moderately importantwas range of products and services. As noted in theliterature, a broader range of products tends toresult in fewer new products being introduced anda more narrow range is associated with greaterproduct innovation. For this reason, the measuredoes seem worthy of the attention of managers,especially in making decisions about the breadthand depth of product lines. The least importantmeasure in the production link measures wasutilization of economic order quantity. It was theonly measure rated less important. It may be thatthe participants, in assigning their ratings, regardedthe use of EOQ as a means to an end ratherthan an end in and of itself. In short, quality andefficiency seem to be more important considerationsin evaluating production performance.5.4. Delivery performance evaluation metricsAfter the orders are planned and goods sourced,produced and assembled, the remaining task is todeliver them to customer. Table 5 shows the orderof importance of delivery performance measures.Quality of delivered goods is first in importance,followed by on time delivery of goods andflexibility of service systems to meet customerneeds. These three measures are highly important.Note that there is very little difference in the ratingof quality of delivered goods and on time deliver ofgoods. Here again, we believe that these three arerelated to the perceived customer value of theproduct, the top ranking strategic planning measure.Providing the customer with a qualityproduct in a timely fashion, and maintainingcustomer satisfaction with a service system designedto flexibly respond to customer needs arekey in producing value for the customer.The effectiveness of the enterprise distributionTable 5Importance of delivery performance measuresAssessmentHighly importantModeratelyimportantLess importantDelivery performancemetricsQuality of deliveredgoodsOn time delivery ofgoodsFlexibility of servicesystems to meetcustomer needsEffectiveness ofenterprise distributionplanning scheduleEffectiveness of deliveryinvoice methodsNumber of faultlessdelivery notes invoicedPercentage of urgentdeliveriesInformation richness incarrying out deliveryPercentage of finishedgoods in transitDelivery reliabilityperformancePercentagerating12.3412.2011.4310.3110.2310.059.328.767.767.70

ARTICLE IN PRESS344A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347planning schedule, effectiveness of delivery invoicemethods, number of faultless delivery notesinvoiced, percentage of urgent deliveries andinformation richness in carrying out the deliveryare moderately important. According to the ratingof measures, while unquestionably important,these measures are not as important as the qualityof the delivered product and on time delivery. Itwould seem, at least on the surface that on timedelivery would result from an effective enterprisedistribution planning schedule, so itwould probably be unwise to ignore the obviousimportance of the enterprise distribution planningschedule—one is the means and the otherthe end.In the survey, companies were asked to expresstheir views on reducing the cost of a deliverysystem. Their responses tended to emphasizetechniques like JIT and the application of automationalternatives to reduce costs. Trade-offsbetween centralisation of the distribution systemand decentralisation of the system were mentionedas was third party logistics.6. A framework for performance measurement ina supplychainIn this section, a framework for performancemeasures and metrics is presented (see Table 6),considering the four major supply chain activities/processes (plan, source, make/assemble, and deliver).These metrics were classified at strategic,tacticaland operationalto clarify the appropriatelevel of management authority and responsibilityfor performance. This framework is based in partof a theoreticalframework discussed by Gunasekaranet al. (2001) and on the empiricalanalysisreported herein. Measures are grouped in cells atthe intersection of the supply chain activity andplanning level. For example, Supplier deliveryperformance can be found at the intersection ofthe Source activity and Tactical planning levelindicating that it pertains to sourcing activities(source) and the tactical planning level. Supplierdelivery performance would thus be a measureuseful in analyzing the performance of mid-levelmanagers as they undertake sourcing activities—mid-level managers who are generally the onesresponsible for tactical decisions.The items in each cell are listed in the order ofimportance based on percentage importance ratings.Those ratings can be seen in Tables 1–5.Readers can refer to those tables in order to moreclosely examine the importance ratings of individualmeasures/metrics.Some measures appear inmore than one cell, indicating that measures maybe appropriate at more than one managementlevel. Measures used at different managementlevels will most assuredly require adjustment totailor them to planning and control needs of thedifferent levels. For example, appropriate measurementmay require that data used by the lowerlevel of management be aggregated in some formor fashion to make the data appropriate for thenext higher level (convert data into informationappropriate for the context). There is nothingnovelabout this approach, as it has been usedfor years in management planning and controlsystems.The approach we used in organizing themeasures for the framework could be used byorganizations in development of a performancemeasurement program for SCM. Managers and/orconsultants could identify measurements (werecommended many such measurements herein),rate their importance using the methodology weused for rating importance, and construct a matrixlike our own to identify the supply chain activity/process to be measured, the measurement, andlevel of management to which the measure shouldbe applied. More detail could be added to fixpersonalresponsibility for measures with individualmanagers,or management positions.Readers should keep in mind that this frameworkis based largely on metrics discussed in theliterature. Individual firms will certainly haveperformance measurement needs that reflect theunique operations of their business and of coursenot all supply chains are identical. Thus othermeasures may be desirable and should be developedby firms and supply chain participants toreflect their unique needs. This framework shouldbe regarded as a starting point for an assessmentof the need for supply chain performance measurement.It is likewise important to understand

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 345Table 6Supply chain performance metrics frameworkSupplychainactivity/processStrategicTacticalOperationalPlanLevel of customer perceived value ofproduct, Variances against budget,Order lead time, Informationprocessing cost, Net profit Vsproductivity ratio, Totalcycle time,Totalcash flow time, Productdevelopment cycle timeCustomer query time, Productdevelopment cycle time, Accuracy offorecasting techniques, Planningprocess cycle time, Order entrymethods, Human resourceproductivityOrder entry methods, Humanresource productivitySourceSupplier delivery performance,supplier leadtime against industrynorm, supplier pricing againstmarket, Efficiency of purchase ordercycle time, Efficiency of cash flowmethod, Supplier booking inproceduresEfficiency of purchase order cycletime, Supplier pricing againstmarketMake/AssembleRange of products and servicesPercentage of defects, Cost peroperation hour, Capacity utilization,Utilization of economic orderquantityPercentage of Defects, Cost peroperation hour, Human resourceproductivity indexDeliverFlexibility of service system to meetcustomer needs, Effectiveness ofenterprise distribution planningscheduleFlexibility of service system to meetcustomer needs, Effectiveness ofenterprise distribution planningschedule, Effectiveness of deliveryinvoice methods, Percentage offinished goods in transit, Deliveryreliability performanceQuality of delivered goods, On timedelivery of goods, Effectiveness ofdelivery invoice methods, Numberof faultless delivery notes invoiced,Percentage of urgent deliveries,Information richness in carrying outdelivery, Delivery reliabilityperformancethat the rated importance of metrics in thisframework is based on a relatively small sample,and thus, care should be taken in generalizingresults to all supply chains. The importance ofindividualmetrics presented herein might notapply to all supply chains in all industries. Again,the framework is only a starting point. It is hopedthat this framework will assist practitioners intheir efforts to assess supply chain performance.7. ConclusionsIn our survey participants were asked whethertheir return on investment had increased toexpected levels after implementing contemporarysupply chain management (SCM) practices. The76% affirmative response to that question clearlyshowed that effort focused on carefully managingsupply chains produced financial benefits forparticipating firms. From a financialperspectivealone, a proactive approach to SCM is advisablefor firms wanting to enhance competitiveness.The SCM literature suggests that effective SCMhelp to win customers and improve customerservice. Some 66% of the respondents in oursurvey noted the positive impact of SCM onmarket share, providing more evidence of thestrategic importance of successfulSCM. Thepotentialbenefits of SCM make it attractive, but

ARTICLE IN PRESS346A. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347improved performance is not automatic. As withany other organisationalundertaking, it must bedone well to yield positive results. This is why webelieve it is important to assess performance inSCM and the reason we developed the SCMperformance measurement framework.To bring about improved performance in asupply chain and move closer to attainment of theillusive goal of supply chain optimization, performancemeasurement and improvement studiesmust be done throughout the supply chain. Allparticipants in the supply chain should be involvedand committed to common goals, such as customersatisfaction throughout the supply chain andenhanced competitiveness. A performance measurementprogram for a supply chain should becomplete—important aspects of performance inany link are not ignored—and they must betailored to varying needs of participants. A goodSCM program will bring about improved crossfunctionalandintra-organisationalprocess planningand control and more complete supply chainintegration. A supply chain wide performancemeasurement initiative would seem most appropriate.This is not to suggest that one party dictatemeasurement programs for all supply chainparticipants, but rather that all participants takepart in developing a well planned, well coordinated,supply chain-wide performance measurementinitiative to which all can and will becommitted. A comprehensive controlsystem wi lbe necessary in order to assure effective andefficient performance measurement all along thesupply chain, but it must not be done in such away as to unduly limit the decision makingauthority of managers in participating organizations.Care must be exercised in developing such asystem in order that it promotes mutually advantageousexchange among participants, so thatrelationships endure the test of time.Additionalresearch and practitioner-driven initiativesare needed in the area of SCM performancemeasurement. Creative efforts are neededto design new measures and new programs forassessing the performance of the supply chain as awhole as well as the performance of eachorganization that is a part of the supply chain.Organisation, suppliers and customers shouldcome together to discuss how they will addressthe measurement and improvement of SCMperformance. Industry consortiums, consultants,and researchers could be helpful in promotingSCM performance measurement generally, and indeveloping measures and measurement techniquesspecifically. They could play a significant role inhelping firms address the present and futurechallenges of managing supply chains. Clearlytremendous opportunity exists to develop measuresthat facilitate progress and promote greatersupply chain integration.AcknowledgementsThe authors gratefully acknowledge the constructiveand helpful comments of two anonymousreferees on the earlier version of the manuscript.ReferencesBallou, R.H., 1992. Business Logistics Management. Prentice-Hall, Inc., Englewood Cliffs, NJ.Benjamin, R., Wigand, R., 1995. Electronic markets and virtualvalue chains on the information superhighway. SloanManagement Review 36 (2), 62–72.Cavinato, J.L., 1992. Total cost value model for supply chaincompetitiveness. Journalof Business Logistics 13(2), 285–291.Christopher, M., 1992. Logistics and Supply Chain Management.Pitman Publishing, London.De Toni, A., Tonchia, S., 2001. Performance measurementsystems: Models, characteristics and measures. InternationalJournalofOperations & Production Management 21(1/2), 46–70.Ellram, L.M., 1991. A managerial guide for the developmentand implementation of purchasing partnerships. InternationalJournalofPurchasing and Materials Management 27(3), 2–8.Fawcett, S.E., 1995. Using strategic assessment to increase thevalue-added capabilities of manufacturing and logistics.Production and Inventory Management Journal36 (2),33–37.Fisher, L.M., 1997. What is the right supply chain for yourproduct? Harvard Business Review 75 (2), 105–116.Graham, T.S., Dougherty, P.J., Dudley, W.N., 1994. The longterm strategic impact of purchasing partnerships. InternationalJournalofPurchasing and Materials Management 30(4), 13–18.Gunasekaran, A., Patel, C., Tirtiroglu, E., 2001. Performancemeasure and metrics in a supply chain environment.

ARTICLE IN PRESSA. Gunasekaran et al. / Int. J. Production Economics 87 (2004) 333–347 347InternationalJournalof Operations & Production Management21 (1/2), 71–87.Kaplan, R.S., Norton, P.D., 1992. The balanced scoreboardmeasuresthat drives performance. Harvard BusinessReview 70 (1), 71–79.Lee, H.L., Billington, C., 1992. Managing supply chaininventory: Pitfalls and opportunities. Sloan ManagementReview 33 (3), 65–73.Levy, D.L., 1997. Lean production in an internationalsupplychain. Sloan Management Review 38 (2), 94–102.Little, D., Kenworthy, J., Jarvis, P., Porter, K., 1995.Scheduling across the supply chain. Logistics InformationManagement 8 (1), 42–48.MacBeth, D.K., Ferguson, N., 1994. Partnership Sourcing: AnIntegrated Supply Chain Management Approach. PitmanPublishing, London.Maloni, M.J., Benton, W.C., 1997. Supply chain partnerships:Opportunities for operations research. European JournalofOperationalResearch 101, 419–429.Mapes, J., New, C., Szwejczewski, M., 1997. Performancetrade-offs in manufacturing plants. InternationalJournalof Operations & Production Management 17 (10),1020–1033.Maskell, B.H., 1991. Performance Measurement for WorldClass Manufacturing. Productivity Press, Inc., Portland,OR.Naim, M.M., 1997. The Book That Changed the World.Manufacturing Engineer (February), 13–16.Novich, N., 1990. Distribution strategy: Are you thinking smallenough? Sloan Management Review 32 (1), 71–77.Rushton, A., Oxley, J., 1989. Handbook of Logistics andDistribution Management. Kogan Page Ltd., London.Schroeder, R.G., John, C.A, Scudder, G.D., 1986. White collarproductivity measurement. Management Decision 24 (5),3–7.Slack, N., Chambers, S., Harland, C., Harrison, A., Johnston,R., 1995. Operations Management. Pitman Publishing,London.Stewart, G., 1995. Supply chain performance benchmarkingstudy reveals keys to supply chain excellence. LogisticsInformation Management 8 (2), 38–44.Thomas, D.J., Griffin, P.M., 1996. Co-ordinated supply chainmanagement. European Journalof OperationalResearch 94(3), 1–15.Trent, R.J., Monczka, R.M., 1994. Effective cross-functionalsourcing teams: Criticalsuccess factors. InternationalJournalof Purchasing and Materials Management 30 (4),3–11.van Hoek, R.I., 2001. The contribution of performancemeasurement to the expansion of third party logisticsalliances in the supply chain. International Journal ofOperations & Production Management 21 (1/2), 15–29.van Hoek, R.I., Harrison, A., Christopher, M., 2001. Measuringagile capabilities in the supply chain. InternationalJournalof Operations & Production Management 21 (1/2),126–147.