Journal of Business Forecasting, Vol.23, Nº4 (Winter 2004-2005)

SALES AND OPERATIONS PLANNING PART II:ENABLING TECHNOLOGYBy Larry Lapide

(This is an ongoing column inThe Journal, which is intended togive a brief view on a potentialtopic of interest to practitioners ofbusiness forecasting. Suggestionson topics that you would like to seecovered should be sent via email [email protected])

This column represents the secondof a three-part series covering theSales and Operations Planning(S&OP) process. As discussed in Part I.S&OP has been receiving a lot ofattention for the past couple of years.There are a number of industry-widestudies in the area.

Companies are recognizing its valuein improving the tactical and operationalplanning to prepare the supply chain formeeting anticipated customer demand.S&OP appears to be driving supply chainbenefits such as better meeting customerdemand while at the same time resultingin reduced inventories and minimizedsupply chain operating costs.

In addition, an indicator of a longerterm interest in the S&OP process is thefact that, according to AMR Research,companies have spent over $12 billion insupply chain planning applicationsoftware over the last 6 years. Yet, whilespending significant sums of money onS&OP-related software, they are notseeing the benefits they expected becausemany did not change the process to fullyleverage the enabling technology.

Hence, the rationale for this column isto see what kind of enabling technology is

needed to support the S&OP process.

THE NEED FORTECHNOLOGY

First and foremost, one needs torecognize that software technology itselfis not very useful. It becomes useful when

LARRY LAPIDE

Dr. Lapide is a Research Director atMIT's Center for Transportation andLogistics where he manages its SupplyChain 2020 Project focused on supplychain management of the future. Hehas extensive business experience inindustry, consulting, and research, andhas a broad range of forecastingexperiences. He was a forecaster inindustry for many years, has ledforecasting-related consulting projectsfor clients across a variety of industries,and has taught forecasting in a collegesetting. In addition, for 7 years he was aleading market analyst in the researchof forecasting and supply chainsoftware.

one starts to improve a business process.However, often, without technology, abusiness process like S&OP iscumbersome and can't support the scaleneeded to achieve all its benefits. In thatcase, technology becomes necessary, butnot sufficient. Often, the process isdealing with a large complex set of needsthat require a level of automation andcomputational sophistication that goesbeyond what can be achieved with manualprocesses merely supported by computerspreadsheets.

Take, for example, the planning needsof a typical Fortune 500 multinationalmanufacturing company. Its S&OPprocess may need to develop weekly plans6 to 18 months out for the followingsupply-demand elements:

Customer demand in termsofStock-Keeping-Units (SKUs) at avariety of shipping locations (i.e.,SKULs). Given that products mightbe shipped from 25 or more plantsand distribution centers fromaround the world, it is not incon-ceivable to have to develop weeklydemand plans for over 100 thou-sand SKULs over a 50-week timeframe or 5 million planningelements.

Finished goods inventory replenish-ment requirements for 25 or moreshipping locations would also resultin millions of numbers of planningelements.

Production plans/schedules withthe corresponding component andmaterial needs of 15 or more plantswould also require millions ofnumbers of planning elements.

Given that there might be a total of

THE JOURNAL OF BUSnMESS FORECASTING, WINTER 2004-05

over 10 million planning elements thatneed to be generated, coupled with thefact that constraint-based planning mightneed to be done using computationallyintensive algorithms, it is often virtuallyimpossible for a Fortune 500manufacturer to support the S&OPprocess manually with spreadsheettechnology alone. This holds true evenwhen the S&OP process is done at ascaled-down aggregated-product level,which is oflen done to make it moremanageable.

ENABLING TECHNOLOGYARCHITECTURE

The S&OP process needs to besupported by three types of softwareapplications: I) demand-side planning, 2)supply-side planning, and 3) an S&OPworkbench. The components and theintegrated supply-demand planningtechnology architecture for these aredisplayed in Figure 1. It depicts how thecomponents need to be integrated amongthemselves, as well as with othertransactional-oriented business systemssuch as Enterprise Resource Planning(ERP), Manufacturing Execution (MES)and Material Requirements Planning(MRP) systetns. The components for eachofthe 3 types of software applications aredescribed below:

/. Demand-Side Planning Systems:These system components support thedevelopment of a demand plan and an"unconstrained" baseline forecast that areused as demand-side inputs to the S&OPprocess. As such, they need to allow usersof the systems to generate statisticalforecasts based on various endogenousand exogenous variables such asMarketing & Sales plans that reflectpromotional campaigns, new productintroductions, pricing actions, and achanging competitive environment, whichhave an impact on future demand. Thenthey incorporate the market intelligenceinto the base line forecasts. The "DemandCollaborator" system, on the other hand,captures, assembles, and processes themarket intelligence gleaned from a variety

FIGURE 1INTEGRATED SUPPLY-DEMAND PLANNING

TECHNOLOGY ARCHITECTURE

S&OP Workbench Dashboards Scorecards

ftDemand-Side Planning

DemandPlanner

DemandCollahorator

t*

f 4Supplv-Side Planning

InventoryMgmt/DRP

Multi-facility APS Inventory Optimizer Supply Collaborator

ERP Systems Legacy Systems MRP Systems Other Transactional MES Systems Systems

of sources such as from field sales andmarketing personnel, as well as fromdownstream customers that share theirdemand forecasting data or are involvedin co-management inventory programssuch as Vendor Managed Inventory (VMI)and Collaborative Planning, Forecasting,and Replenishment (CPFR). To facilitateinformation collection from remotelocations and external sources such ascustomers, a "Demand Collaborator" isusually web-based so that the Internet canbe leveraged to transfer informationaround.

2. Supply-Side Planning Systems: Thesesystem components support thedevelopment of supply plans that are usedas the supply-side inputs to the S&OPprocess. As such, they help to generate theinventory, production, and procurementplans that will be followed to best meetthe "unconstrained" baseline demandforecasts. These plans might result insupporting a "constrained" demand

forecast when supply capacity isinsufficient to meet all expected customerdemand. The role of 'inventory Manage-ment and Distributed RequirementsPlanning (DRP)" systems is to supportusers in coming up with the expectedinventory replenishment needs of finishedgoods warehouses, such as customer-facing warehouses and the centralizedwarehouses that might replenish them. Inconstrained supply environments, "Multi-facility Advanced Planning andScheduling (APS)'" systems are used toproduce more accurate plans that accountfor limitations in plant and distributioncapacity, as well as for any short-supplyof components, materials, and otherproduction resources. "InventoryOptimizer" systems support these types ofsystems by helping users set inventorytargets during the S&OP process thatoptimally tradeoff customer servicetargets against component, material, sub-assembly, and fmished goods inventories.Lastly, "Supply Collaborator" systems are

THE JOURJNJAL OF BUSINESS FORXCASTING, 200405 19

used to capture, assemble and processsupply capabilities from a variety ofsources such as from purchasingpersonnel and upstream suppliers,including contract manufacturers. Tofacilitate information collection fromremote locations and external sources, a"Supply Collaborator" is usually web-based so that the Internet can be leveragedto transfer information around.

3. S&OP Workbench: This systemcomponent supports two types ofinformation needed to be shared duringcross-functional S&OP meetings. First,using the workbench, generate dash-boards to display a multitude of metricsthat portray the planned supply versus"unconstrained" demand situation. Theseinclude supply-side metrics like expectedplant utilizations, production capacityshortages, and critical componentshortages/surpluses; as well as demand-side metrics such as expected unfulfilledcustomer demand and expected customerorder backlogs. The dashboard function-ality also allows S&OP participants toquickly conduct what-if analyses ofpotential changes to the supply and/or thedemand plans. (Since a complete regene-ration of all supply-demand plans usuallytakes too long, the what-if analyses cannotinclude a full run of all the supply anddemand systems, and an incrementalapproach is usually necessary. Forexample, one Semi-conductor company Italked to said it took 12-hours to run theirsupply planning systems certainly toolong to support these types of what-ifanalyses during an S&OP meeting.) Asecond type information that is neededduring the S&OP process is how well theprocess itself is working. Thus, using theworkbench, scorecards of Key Pertbrm-ance Indicators (KPIs) that reflect howwell the S&OP process has been workingfoster learning and improvement to theprocess over time. Some of these KPIsinclude metrics such as demand forecastaccuracy, variance to the baselineforecast, and adherence to both the supplyand demand plans previously put in place.

As shown in Figure 1, the demand-

side and supply-side planning systemsneed to be integrated and synchronized sothat a change in either the demand or thesupply plan can be quickly reflected in theoverall supply-demand picture. TheS&OP workbench also needs to beintegrated and synchronized with theseand other planning components so thatany changes made in plans during orbetween S&OP meetings can be reflectedin the workbench's supply-demandpicture, as well.

OFF-THE-SHELFTECHNOLOGY

A large part of the technologyarchitecture described above is availableas off-the-shelf supply chain planningapplications in the market today. Much ofit primarily the Demand-Side andSupply-Side planning components aremarketed by the major ERP softwarevendors that have bolstered their SupplyChain Management (SCM) applicationsover the past 5 years, as well as from thespecialty SCM software vendors that areresponsible for the early innovations indeveloping and marketing supply chainplanning software.

Off-the-shelf S&OP workbenchfunctionality, however, is less available. Afew ERP and SCM software vendors arebeginning to offer some of the off-the-shelf functionality needed. BusinessIntelligence (BI) software vendorsspecialize in software that can createmuch of the dashboard and scorecardcapabilities that are needed as part of anS&OP workbench, but do not generallyoffer (off-the-shelf) Demand-Side andSupply-Side planning components.

Do you need all the components ofthe S&OP technology architecturediscussed above? Probably not. In mythird and last column of this S&OP series,I will offer a diagnostic toot that can beused to help you improve your S&OPprocess by assessing what stage of an"S&OP Maturity Model" your company'sprocess is currently at, as well as theprocess changes that may need to be made

mt jOUKNAL oh BUSINESS i-ORECASTlNG

SUBSCRIBE TOTHE JOURNAL OF

BUSINESS FORECASTING

A host of jargon-free articles on howto obtain, recognize, and use goodforecasts written in an easy-to-understand style for businessexecutives and managers. Plus, it willgive new, practical forecasting ideas tohelp you make vital decisionsaffecting sales, capital outlays, finan-cial planning, budgeting, productionscheduling, marketing strategies andfinancial planning.

The Consensus Forecasts consists of19 separate forecasts of 13 keybusiness and economic indicators plusa consensus. The list of participants inthe forecasts reads like Who's Who offinancial markets.

International Economic Outlook givesone year ahead forecasts of realGNP/GDP growth rate, inflation andcurrent aecount of the balance ofpayments of 47 different countriesaround the world. Plus, it givesquarterly forecasts of interest rates offive major markets and quarterlyexchange rate forecasts of 8 majorcountries - Japan, Germany, England,Switzerland, France, Italy, Canada andAustralia.

Subscription Form

(Published four times a year)Subscription {hard copy): $85 Domestic,

$110 Foreign including Canada(PDF file): $55 Both Domestic and Foreign

The Journal of Business Forecasting350 Northern Blvd., Suite 203

Great Neck, N.Y. II021516.504.7576 800.440.0499

Email, [email protected] Web. www.ibf.org

to improve it. And it is these processchanges that will dictate the types ofenabling technology you would need toput in place. Remember, businessprocesses dictate the enabling techno-logies that one needs for supply chainimprovement!

20 THE JOURNAL OF BUSINESS FORECASTING, WINTER 200405

Rafa CarranzaResaltado