欧莱雅集团供应链诊断报告课件.ppt

1、2022年3月25日星期五欧莱雅集团供应链诊断欧莱雅集团供应链诊断报告报告Presentation Agenda I.Project Scope and ActivitiesII.Overview of Our Supply Chain FindingsI.ProcessesII.Inventory ReviewIII.Maturity ProfilesIV.Organization/ People and Culture V.TechnologyIII.Supply Chain RoadmapIV.Discussion of Next StepsOur Project ApproachOve

2、r the past three weeks, we have:Interviewed 5 Executives of LOral CanadaFacilitated two cross-functional process workshopsCollected and analyzed dataConducted a high level maturity assessmentCompared LOral process and performance to CPG best practiceIdentified key business issues and areas of improv

3、ementSummarized our findings in this presentation documentProject KickoffData CollectionInterview with ExecutivesProcess WorkshopsModel Review & Validate Base CaseReview MetricsIdentify & Consolidate Issues & OpportunitiesOpportunity PrioritizationBenefits AssessmentFindings/ Recommendation Developm

4、entPresent RecommendationsDiscuss Next StepsCross-Functional Participants Laurent Venot (Project Sponsor)Jean ValoisPierre MassicotteValry MayerDominique De Celles Isabelle Genest Anne-Marie NelsonDion GemmityEli BengioPatricia RaposoValrie LacerteYves DubNaity JacelPierre Journel Marcelyne CraftMar

5、tin DeschnesPresentation Agenda I.Project Scope and ActivitiesII.Overview of Our Supply Chain FindingsI.ProcessesII.Inventory ReviewIII.Maturity ProfilesIV.Organization/ People and Culture V.TechnologyIII.Supply Chain RoadmapIV.Discussion of Next StepsProcesses Develop Sales Forecast Maintain Promot

6、ional Plan Consider inventory levels, backorder reports, safety stock levels Consolidate and consider external data (POS, Nielsen, Customer data) Reconcile Demand Plan with Budget New Product Introductions Net Inventory and Lead Times from the Forecast Distributed Requirement Planning Capacity Requi

7、rement Planning Customer Allocations Communicate forecast to Suppliers Receive supplier commitment Process Customer Orders Plan and Manage Finished Goods Inventory Picking and Packing Manage Outbound Product Flow Manage shipping & transportation Generate Invoice Credit management and A/R(*) Out of s

8、copeDemand Planning Issues and OpportunitiesIssuesManual and Inconsistent ProcessesThe forecast process is mostly manual, inconsistent across brands and divisions (due to different levels of maturity) as well as both time and resource consumingThere is no formal process for determining which product

9、s should be focused on during the forecasting meetingsNo one department “owns the forecast”There is no mechanism for automatically generating a baseline forecastToo much time is spent on manual processes, data clean-up and generating reports, leaving insufficient time to analyze data and manage exce

10、ptionsLOral Canadas brand forecasts are not all aggregated up into one forecast limiting visibility and data sharing LOral Canada does not have simulation or “what if” capabilitiesMany disparate tools are being used to facilitate forecasting - Excel, Valo, SAP and Outils Gestion. Interfaces have bee

11、n built between the various tools and systems, but manual verification is required and the current technology still does not meet LOral s requirementsPOS, Nielsen, customer information, promotional and advertising information are not all integrated in one central repositoryOpportunity Process Redesi

12、gn Spend time forecasting products that give the highest return. Not all product lines benefit from increased attention. Focus on high margin, low volume, big ticket items, high profile items and items with complements to maximize profitability Communicate and reconcile forecast assumptions by condu

13、cting a formal review after each planning cycle with all stakeholders. Report forecast error and modify assumptions as appropriate Forecasting Tool/BW Implementing a Forecasting tool will enable the automatic generation of a baseline forecast based on sales history and additional variables. Most pac

14、kages allow users to select a statistical model, enable the system to determine the model with the greatest fit or employ a composite model The forecasting tool selected should allow for the generation of alerts based on predefined tolerances, highlighting the products that require further considera

15、tion and manual intervention Products should be reviewed to determine which ones are critical or high margin to plan in a Forecasting tool, while non-critical or stagnant products may be planned in SAP R/3 The forecasting tool selected should provide “What If” simulation capabilities to demonstrate

16、the impact of a variable change (e.g. price change) Coordinate data collection for all available sources and house it in a central repository such as SAP Business WarehouseDemand Planning Issues and OpportunitiesIssuesOpportunityPeopleEnsure the Planning Team is Cross-Functional and Cross-Trained: L

17、Oral will need to evaluate their resources skills and determine if additional resources are required to gain forecasting expertise Statistical proficiency and system expertise should be balanced with industry, company and product knowledge Cross-training sharpens skills, broadens perspectives and im

18、proves communication Create incentives to improve the accuracy of forecasts: Establish forecast accuracy goals Establish compensation structure Monitor and review congruence of goals to rewards This will remove trade-offs between forecast accuracy and performance against forecast, leverages key info

19、rmation and helps eliminate biasesLOral Canada has started to tie compensation to forecast accuracy. This practice will obtain greater acceptance among employees once a tool to support the process has been implementedDepartments Involved have Different PrioritiesWhile the generation of the forecast

20、is a collaborative effort between Sales, Marketing, Finance and Logistics, the involved parties do not share the same view of the forecasts importance nor appreciate the impact of their input (or lack thereof) on other departments ability to execute their functions downstreamSales is seen as too dis

21、tant from the customer and therefore lacking pertinent data to input into the forecasting e.g. which store carries which SKUSales strategy to push products at the end of the month and year by decreasing prices, throws off the forecastLOral Canada realizes that even if they implement a forecasting to

22、ol, without the proper processes or resources in place, forecast accuracy will not likely improve LOral Canada experiences a high turnover among sales and marketing people resulting in the loss of vital customer and product informationTraining for Sales and Marketing resources is insufficient. Compe

23、nsation re-evaluation has begun to align individual goals and metrics with organizational goals with respect to forecast accuracyAlthough LOral Canada has made great strides in improving their forecasting process, the accuracy has not improved significantly leading to feelings of frustration among e

24、mployeesDemand Planning Issues and OpportunitiesIssuesOpportunityProduct Launches Although LOral Canada has access to other countries launches as well as to like products sales history, there is no tool or process in place to effectively incorporate this data into their forecastsLOral Canada is unab

25、le to accurately assess the cannibalization effect of one product on anotherNo post-mortem analysis is conducted or captured on failed product launches as input to future launchesPromotionsThere is currently no means to accurately cleanse promotional uplifts from sales history Past promotional perfo

26、rmance is not adequately measured, therefore making it difficult to predict the impact of future promotionsPromotions involving giftware can lead to imbalances in demand, making it difficult to predict true demand for these products“One Shot” promotions are difficult to planProduct AllocationProduct

27、 allocation has not been deployed effectively and has even led to undesirable forecasting behaviours Product Lifecycle Management Investigate APS functionality to handle: “Like Modeling” whereby sales history of a similar product is used to forecast a new product Phase In/ Phase Out of products whic

28、h entails weighting the forecast to reflect a products lifecycle The separation of the promotional uplift from baseline forecast generation CannibalizationData Cleansing Sufficient time must be spent cleaning data such as SKU changes, data errors, duplication, one-time events, stock-outs, promotions

29、, outliers etc to enable forecasting models to separate business growth from seasonality and other deviationsProduct Allocation Allocation is a useful tool for ensure higher priority customersobtain increased service levels in cases of shortages. LOral should develop and implement formal policies fo

30、r allocating products (e.g. based on customer forecasts) Allocation configuration should be modified to reflect the new policies in SAP R/3. In the long-term additional allocation functionality could be explored in an APS toolDemand Planning Issues and OpportunitiesIssuesOpportunityCustomer Collabor

31、ation Cross-Share Forecasts with Key Customers Incorporate customer forecasts which are deemed as reliable into baseline forecasts and challenge their numbers (manage by exception) Once forecasts are more accurate, LOral should provide a website to allow easy access for their Customers to enter thei

32、r forecasts and to post LOrals confirmed quantities for customers to validate in a timely manner Aggregation / Disaggregation Determine the appropriate level to forecast e.g. SKU/customer location for DSD, SKU/Customer DC for warehouse shipments An APS solution will facilitate the aggregation and di

33、saggregation of the forecast, however, time should be spent determining the appropriate level for forecasting. Top-down approaches can mask trend shifts while bottom-up approaches may not consider effects shared across brands Measuring Forecast Inaccuracy Order date should be used as the point for e

34、stablishing demand since it gives the most accurate information on when demand is initiated. Otherwise stock-outs, lost sales, opportunity costs and lost customers may be masked Ensure the full aggregated costs of forecast error are measured: Insufficient inventory (stock outs) Excess inventory (ret

35、urns, spoiled, destroyed, obsolete, slow moving) Safety stock Expediting costs Inventory carrying costs OvertimeCustomer CollaborationAlthough some of LOral Canadas customers provide forecast and POS data, LOral does not possess the infrastructure or processes to effectively use the dataLOral Canada

36、 must first focus on improving its own processes before extending their processes and data out to share with Customers and SuppliersForecast Accuracy While forecast accuracy is better at the Brand level, it varies greatly at the SKU level regardless of product classification (A, B, C) Only 40% to 50

37、% of items achieve the target of tolerance of 20% (BIC 75% of SKU achieve this target) SKU ranking has been erroneousImpacts of Forecast InaccuracyNumerous forecast modification requests to the factoriesCustomer Service levels (line fill rate in $) are at 92% despite target s of 98.5% (BIC is 99%)In

38、ventory Write-offs are at 10.5% (BIC for CPG is 2%) Inventory Turns are 4 (BIC for CPG is 17) Sales Returns are at 5.25% (BIC is 0.4%) NOTE: We are comparing LOral to CPG companies which is not a fair comparison in some cases. The comparison is meant as a guideline.Supply Planning Issues and Opportu

39、nitiesIssues Input The unconstrained forecast is the main input for developing the purchasing plan. Given LOral Canadas poor forecast accuracy, the purchasing plan does not effectively meet customers demands as demonstrated by the numerous forecast modification orders sent to the plants and high inv

40、entory levels Products lead times, safety stock levels, min/max levels are not reviewed on a regular basis Factory Constraints The purchasing frozen horizons set by the plants seem excessively long and are not reviewed sufficiently. Therefore, LOral Canada must place orders far in advance, often bef

41、ore they are able to assess trends The plants confirm back all ordered quantities since replenishment lead time is used. As a result, LOral does not obtain a real view of supply or capacity constraints The factories are Product focused rather than Sales/Customer focused. The result is a Push model r

42、ather than Pull. LOrals factories occasionally push excess stock to the DCs rather than trying to match customers requirements No global supply planning is engaged in at LOral. Each LOral Sales Company is prescribed where to buy each product and can only look to other factories in cases of shortage

43、/ capacity constraintsOpportunity Factory Constraints Collaborative Planning with the plants would involve LOral Canada posting their forecast on the Web and allowing the factories to confirm the quantities they can supply against the forecast in a timely manner Continue to roll-out one of the newer

44、 LOral Canada processes whereby products are stocked in the US and shipped directly to customers in Canada to reduce inventory levels and costs Long Term Once LOral Canada has improved their forecast accuracy, they will be in a better position to approach the factories to discuss: Shortening frozen

45、horizons Communicating issues Obtaining visibility to the plants production schedules to gain a better sense of supply plans and capacity constraints Receiving ASNs to give LOral Canada a true sense of their ATP to pass on to their customers Develop new policies and education program across LOral co

46、mpanies to ensure customer service levels are the primary focus rather than factory throughput and costs All LOral company forecasts should be rolled up to a central supply chain group for global supply planning purposes. This would enable global fulfillment to match supply with demandSupply Plannin

47、g Issues and OpportunitiesIssuesOpportunityAllocation The US Factories ought to be allocating stock in ratio to LOral Sales Companies forecasts. This would:Encourage the LOral Sales Companies to improve their forecast accuracyEnsure fair share allocation in shortage situations since each Sales Compa

48、ny would obtain stock in proportion to their forecastBeyond the forecasted amount stock could be allocated on a “first come, first serve” basis The US Factories should measure and be compensated based on their ability to service all of North America and not just the US VMI Implement an APS tool that

49、 includes VMI functionality and choose a pilot customer to model with before extending the functionality out to other customers Leverage the VMI experience of LOral USA (e.g. Lancme)Product Harmonization Where possible, continue to roll out the process of manufacturing with Final Assembly, whereby p

50、roducts are produced and finished to the point of localization steps Factory Constraints (contd) Canada makes up only 3% of LOrals global sales, while the US which supplies Canada makes up 28% of LOrals global sales. Therefore, LOral Canadas requirements are only fulfilled once the USs needs have be