Inventory Kimball & Ross, Chapter 3OverviewValue ChainInventory ModelsInventory Periodic Snapshot ModelInventory Periodic Snapshot Model - ChallengeSemiadditive FactsEnhanced Inventory FactsInventory Transactions ModelInventory Transactions Model - Con’tInventory Accumulating Snapshot ModelValue Chain IntegrationData Warehouse Bus ArchitectureData Warehouse Bus Architecture – Cont’dData Warehouse Bus MatrixConformed DimensionsTypes of Dimension ConformityCentralized Dimension AuthorityConformed FactsAcknowledgementsInventoryKimball & Ross, Chapter 3OverviewValue chain implicationsInventory periodic snapshot model, transaction and accumulating snapshot modelsSemi-additive factsEnhanced inventory factsData Warehouse bus architecture and matrixConformed dimensions and factsValue ChainThe value chain identifies the natural, logical flow of an organization’s primary activities. See Fig. 3.1Operational source systems produce transactions or snapshots at each step in the value chain. They generate interesting performance metrics along the way.Each business process generates one or more fact tables.Inventory ModelsInventory periodic snapshotInventory level of each product measured daily (or weekly) – represented as a separate row in a fact tableInventory transactionsAs products move through the warehouse, all transactions with impact on inventory levels are recordedInventory accumulating snapshotOne fact table row for each product updated as the product moves through the warehouseInventory Periodic Snapshot ModelBusiness needAnalysis of daily quantity-on-hand inventory levels by product and storeBusiness processRetail store inventoryGranularityDaily inventory by product at each storeDimensionsDate, product, storeFactQuantity on handInventory Periodic Snapshot Model - ChallengeVery dense (huge) fact tableAs opposed to retail sales, which was sparse because only about 10% of products sell each day60,000 items in 100 stores = 6,000,000 rowsIf 14 bytes per row: 84MB per dayOne-year period: 365 x 84MB = 30GBSolution: Reduce snapshot frequencies over timeLast 60 days at daily levelWeekly snapshots for historical dataFor a 3-year period =208 snapshots vs. 3x365=1095 snapshots; reduction by a factor of 5Semiadditive FactsInventory levels (quantity on hand) are additive across products or stores, but NOT across dates = semi-additive factsCompare to Retail Sales:once the product is sold it is not counted againStatic level measurements (inventory, balances…) are not additive across date dimension; to aggregate over time use average over number of time periods.Enhanced Inventory FactsNumber of turns = total quantity sold / daily average quantity on handDays’ supply = final quantity on hand / average quantity soldGross profit = value at latest selling price - value at costGross margin = gross profit / value at latest selling priceGMROI (Gross Margin Return On Inventory) GMROI = number of turns * gross marginmeasures effectiveness of inventory investmenthigh = lot of turns and more profit, low = low turns and low profitNeed additional facts: quantity sold, value at cost, value at latest selling price GMROI is not additive and, therefore, is not stored in enhanced fact table. It is calculated from the constituent columns.Inventory Transactions ModelRecord every transaction that affects inventoryReceive productPlace product into inspection holdRelease product from inspection holdReturn product to vendor due to inspection failurePlace product in binAuthorize product for salePick product from binPackage product for shipmentShip product to customerReceive product from customerReturn product to inventory from customer returnRemove product from inventoryInventory Transactions Model - Con’tDimensions: date, warehouse, product, vendor, inventory transaction type.The transaction-level fact table contains the most detailed information possible about the inventory.It is useful for measuring the frequency and timing of specific transaction types.It is impractical for broad data warehouse questions that span dates or products. To give a more cumulative view of a process, some form of snapshot table often accompanies a transaction fact table.Inventory Accumulating Snapshot ModelBuild one record in the fact table for each product delivery to the warehouseTrack disposition of a product until it leaves the warehouseReceivingInspectionBin placementAuthorization to sellPickingBoxingShippingThe philosophy of the inventory accumulating snapshot fact table is to provide an updated status of the product shipment as it moves through above milestones.Rarely used in long-running, continuously replenished inventory processes.More on this in chapter 5.Value Chain IntegrationBoth business and IT organizations are interested in value chain integrationDesire to look across the business to better evaluate overall performanceData marts may correspond to different business processesNeed to look consistently at dimensions shared between business processesNeed an integrated data warehouse architectureIf dimension table attributes in various marts are identical, each mart is queried separately; the results are then outer-joined based on a common dimension attribute = drill acrossData Warehouse Bus ArchitectureCannot built the enterprise data warehouse in one step.Building isolated pieces will defeat consistency goal.Need an architected incremental approach  data warehouse bus architecture.See Fig. 3.7By defining a standard bus interface for the data warehouse environment, separate data marts can be implemented by different groups at different times. The separate data marts can be plugged together and usefully coexist if they adhere to the standard.Data Warehouse Bus Architecture – Cont’dDuring architecture phase, team designs a master suite of standardized dimensions and facts that have uniform interpretation across the enterprise.Separate data marts are then developed adhering to this architecture.Data Warehouse Bus MatrixSee Figure 3.8The rows of the bus matrix correspond to business processes  data martsSeparate rows should be created if:the sources are different,the processes are different, ora row represents more