11 Calculation Concepts

This chapter provides conceptual information about the SRP and WRP calculations. Formulas for these calculations are available in the My Oracle Support Note: Calculations for Store and Warehouse Replenishment Planning 1452505.1.

Replenishment

Replenishment is an AIP batch process that calculates the replenishment plan for each SKU Pack-size at each location across the planning horizon. Replenishment assumes no quantity constraints and so the calculated orders reflect the location's need without regard for the ability of the source of the orders (warehouse or supplier) to supply them.

However, real world constraints upon the supply of product do exist. This is addressed by splitting the planning horizon for each SKU Pack-size/location into two distinct periods of time. The first period contains that part of the planning horizon over which constraints apply. This is referred to as the fixed period because the ultimate supply of stock within this time period is limited, or fixed. The second period of time is called Post Fixed Period and refers to that remaining part of the planning horizon that exists after the fixed period.

A special case of replenishment is when a SKU is capped. The Inventory Cap represents a physical space constraint in units at a store for a given day. The Inventory Capping functionality generates a Store replenishment plan such that inventory (including planned receipts, expected receipts, and current inventory) at the store does not violate the space constraints at the store for any day. At the same time, the Warehouse replenishment plans are generated based on forecasted demand from the stores that have no inventory caps being applied to it across time.

Replenishment is run three times each night—the first time it produces an Unconstrained Receipt Plan (URP) for the entire planning horizon for all valid SKU Pack/locations. This is done because some replenishment methods at the warehouse level require demand over a specified period of time that may exceed the fixed period. The second time it is run only for the fixed period to establish the capped URPs for SKU/stores with inventory caps applied. Finally, it is re-run post fixed period to adjust the plan to inventory constraints that occurred within the fixed period. In between the second and third replenishment runs, the constraints that relate to the fixed period for each SKU Pack/location are applied. These constraints modify the plan within the fixed period and ultimately determine the correct inventory position at the start of the post fixed period. This updated inventory position is then used by the third run of the replenishment process.

Replenishment looks at the current inventory position, forecasted demand, customer orders, and expected receipts to project the inventory position in the future. Based on the user-defined Replenishment Method and other user-defined parameters, stocking level targets (allocation boundaries) for a product/location combination are calculated and used in generating an ideal receipt plan that is not limited by available inventory constraints.

Intra-day Replenishment

Intra-day Replenishment performs similar activities as overnight replenishment but over a more limited order set and time horizon. This Intra-day process can be performed many times during the day.

The purpose of the intra-day activities is to determine recalculated order quantities using updated dynamic data feeds such as sales and inventory, for those delivery opportunities that are to be executed today on an intra-day release wave.

Intra-day activities are limited to those required to process into-store orders. Therefore, only store orders can reap the benefits of updated inventory and sales feeds.

Into-store orders are released throughout the day at discreet times called Waves. Each into-store order has the opportunity to be recalculated immediately prior to its release (if you choose). If you choose not to recalculate an order, then the quantity released is the one generated by the overnight Replenishment action. If you choose to recalculate into-store orders immediately prior to their release, then intra-day replenishment will identify the SKU/Store/Days to be re-planned, revise sales forecasts as needed, perform replenishment on the limited set of SKU/Store/Days, and perform reconciliation of the limited set of sources that supply the SKU/Store/Days being re-planned.

The following steps are unique to Intra-day Replenishment:

Identify What Plans to Recalculate
Revise the Sales Forecast

Identify What Plans to Recalculate

It is necessary for the intra-day activities to be aware of which wave is being performed. This is because each run of the intra-day process deals with a much smaller cross-section of the supply chain network. To properly identify which order plans should be recalculated on a particular intra-day wave, the following steps must be performed:

Identify both warehouse and vendor sources that serve the SKU/Store/ATP day combinations that meet all of the following criteria:
- The SKU/Store is assigned to the current wave for the current day and the user has indicated the SKU/Store should be re-planned prior to release.
- The SKU/Store has an ATP day whose release date is today.
- The SKU/Store is not inventory capped.
- An assumption is made that related Event SKUs are assigned to the same wave and have the same re-plan indicator. This is required so that both SKUs can be planned together for the purposes of subbing and combining inventory. The long-term SKU's wave is applied to the temporary SKU (Promotional Varient or SKU going through replacement).
Note:
Because only Today will be reconciled no vendor/SKUs with an applicable SPQ will be part of Intra-day Replenishment.

Using the previous list of SKU/sources identify the list of all SKU/Store/ATP day combinations that meet all of the following criteria:

The SKU/Store is assigned to the current or later wave for the current day and you have indicated the SKU/Store should be re-planned prior to release.

The SKU/Store has an ATP day whose release date is today and the source for the SKU at the store for that ATP day is a SKU/Source identified in the previous step.

Note:

Some of these orders may not need to be released until later waves, but by considering them in the current run, the Reconciliation process has greater flexibility to distribute inventory optimally.

Revise the Sales Forecast

Intra-day processing takes place during the day and by this time there may be Actual Sales reported for some SKU/Stores for the first day in the planning horizon. If Actual Sales are reported, then it is possible and maybe necessary to revise the forecast demand for the first day in the planning horizon. Such situations can arise where the Actual Sales for the current day already exceeds the forecast demand or represents a value that indicates that the existing forecast demand for the current day in inaccurate.

Note:

Intra-day processing can take place many times a day. It will only revise the forecasts of those SKU/Stores to be processed on the current wave.

There are two approaches to revising the forecast during the Intra-day Replenishment batch. A User-controlled parameter (Intra-day Revise Sales Forecast Method) indicates which approach should be adopted.

The first approach is Set to Actual. This addresses situations where today's Actual Sales of a SKU at a Store already exceeds the forecast demand for that day. In these situations, the forecast demand for the day are set to the same value as the Actual Sales.

The second approach is referred to as Recalc. If a Sales Profile by hour is available, along with a time that the Actual Sales were captured, then it is possible to make a more intelligent adjustment to the forecast demand for the day. This more intelligent adjustment takes a more conservative approach. It assumes that whatever Actual Sales have already taken place, the remaining Sales for the day will be in line with the original Sales Forecast used overnight. In effect, the new total is the sum of the Actual Sales up until the time the sales were polled, plus the Original Forecast Demand for the remaining part of the day. Note that Original Forecast Demand is always the one used in the previous over night batch run - not any revised (new) forecast produced by a previous intra-day run.

Replenishment Calculation Variables

The following sections describe replenishment calculation variables.

Fixed Period

The term Fixed Period refers to the period of time for a location inside which the supply of inventory for a SKU Pack is ultimately constrained, or fixed.

In a typical multi-tier supply chain the fixed period is the time period it takes for an order to be delivered from a supplier, and pass through the appropriate levels of the supply chain to reach the destination whose fixed period is being calculated. The time it takes a product to be moved from one location to another is the lead-time. As a result the fixed period for a typical SKU Pack/location is the sum of the lead times plus any days from the point of receipt to the point of shipment, and with consideration for Delivery-day Demand Percent, for all locations an order must pass through from Supplier to the location.

This is true when the Supplier has infinite availability of supply and therefore no constrained, or fixed days. Typically AIP assumes infinite availability of supply from a Supplier. The one exception is those Suppliers that are subject to a Supplier Purchase Quantity (SPQ) agreement. The presence of any SPQ arrangement inside the planning horizon should result in the FP for that Supplier being set to the number of days commensurate with the last day of the week of the last week containing an SPQ. The calculation of FP for nodes of the supply chain below the supplier needs to find the first delivery day in the planning horizon on or after the point at which inventory from the source is unconstrained and able to meet demand.

Available To Plan (ATP) Dates

The following sections describe ATP Dates.

Release Schedules

A lead-time is the number of days ahead of the desired delivery date than an order must be communicated to the entity (or source) responsible for making that delivery. Sources for a delivery into a store can be vendors, or warehouses. Sources for a delivery into a warehouse can be vendors, or other warehouses. Lead-times are ultimately Source/SKU Pack/Destination/Delivery Date specific. This means that a lead-time on a specific date indicates the advanced notice that the source of a specific SKU Pack requires to make a delivery into the destination on that date.

A release schedule holds a lead time for each Source/SKU Pack/Destination/Delivery date within the planning horizon that is able to plan a delivery on that day. A release schedule therefore indicates whether a delivery can be received at a location and if so, the lead time required for receiving it.

Planning Deliveries

A delivery cannot be planned on a date whose lead time would require ordering in the past. Imagine today is Monday and the lead-time for delivery into a warehouse on every Tuesday is two days. The release schedule must indicate that tomorrow is not a legitimate day. This is because the lead time for tomorrow would be two days, indicating that a delivery tomorrow must have been planned and executed yesterday. There may already be an order expected for delivery tomorrow, but you cannot re-plan the delivery today because the opportunity for that passed yesterday.

Primary and Secondary Schedules

The batch module is responsible for producing schedules and actually produces two different schedules for use by the AIP replenishment engine. These are Primary and Secondary Schedules.

Primary Schedules are used by Replenishment to determine the days on which orders can be planned into a store or warehouse destination. Secondary Schedules are used in Reconciliation only when a warehouse experiences a shortage at which point the Reconciliation logic may turn to a Secondary (or alternate) source to address the shortage.

For stores, the Primary Schedule is constructed only for the single source that supplies the SKU to the store on that day. For warehouses, the Primary Schedule is constructed using the Source Split information. Source splits indicate what percentage of a warehouse's order should be given to one or more sources that are able to deliver to that warehouse. While many sources may be able to deliver a product to a warehouse, the user may decide to attribute a percentage of the order to only one or a few of the sources. For a delivery day to be a legitimate day in the release schedule, a percentage must be assigned to the source for the warehouse destination and SKU Pack. The lead time schedule for any source not designated as satisfying a percentage of an order on a particular day into a warehouse must be blank on that day. Sources that have been attributed a percentage of a warehouse's need and therefore appear in the Primary Schedules are considered to be Primary Sources.

The Secondary Schedule does not observe the source splits, but instead considers the list of secondary (alternate) warehouse sources specified by the user. Secondary Schedules is only built for those sources that have been specifically identified as secondary sources for a given warehouse by the user.

ATP Days

ATP days are those days on which Replenishment performs an order calculation to determine if an order is needed. ATP days are determined by examining the Primary Release Schedule for each source (bearing in mind there could be more than one) that can deliver a product into a destination on a given day. If any one of these sources can deliver into the destination on a day, then that day is considered an ATP day. This is because the day represents a potential delivery opportunity into the destination and therefore a planning calculation should be performed for that Destination/Product/Day combination in order to determine what, if any, quantity, is needed.

Alternates as Primaries Functionality

The Alternates as Primaries functionality allows you to change the way in which the Primary Schedule is built. This functionality specifically allows Secondary (or alternate) sources to produce Primary Receiving Schedules, even though the Source does not meet the normal qualifying Source Split criteria required to produce a Primary Schedule. This functionality addresses those situations where there are multiple sources that can deliver into a warehouse destination, but the favored source (presumably due to cost reasons) has a longer lead time.

A Tale of Two Vendors

Imagine a situation where there are two vendors—one considerably more expensive to purchase from than the other. The cheaper vendor is given one hundred percent of all the orders to satisfy into a warehouse (the Source Split is 100% for this vendor). However, this cheaper vendor also has a lead time of 10 days. This means that the Release Schedule for this vendor is blank in the first nine days of the planning horizon because any order to be delivered in the next nine days must have been executed before today. Therefore, there are no ATP days at the warehouse in the first nine days. The expensive vendor has a lead time of only three days. Traditionally, the user is forced into a decision between choosing a cheaper, less-reactive vendor, who can delivery no earlier than 10 days from now, and a more expensive vendor who can deliver in only three days time. The Alternates as Primaries indicator addresses this choice.

In its off state of zero (0), the indicator maintains the already described behavior when building the Primary Schedules. This means that the expensive vendor with a source split is considered a Primary Vendor and the Primary Schedule contains only this vendor but the first delivery in the Primary Schedule is in 10 days time. Days one to nine are not considered valid delivery days and therefore replenishment does not consider days one to nine as ATP days.

If the indicator is set to a non-zero state, then the behavior that builds the Primary Release Schedules considers the more expensive vendor if it has been set up as a secondary vendor. In this situation, the secondary (or alternate) vendor is considered a Primary Vendor for the one or more days in which it can deliver before the first real Primary Vendor can deliver. The real Primary Vendor can only deliver from day 10 onwards. The Secondary Vendor can deliver from day three onwards. The functionality can therefore consider the Secondary Vendor to be Primary between days three and nine inclusive. If the indicator is set to one, the Secondary Vendor is only considered to be Primary on the 1st day (day three) that it can deliver before the true Primary can deliver (day 10). If the Indicator is set to two, then the Secondary Vendor is considered as Primary for all days it can deliver before the real Primary can deliver (days three to nine inclusive).

Secondary Sources as Primary Sources Impact

The net result is that the Secondary Source is considered a Primary Source for one or more days before the real Primary Source can deliver, and is written out to the Primary Receiving Schedule. The impact of this is that either day three (indicator set to one) or all days between days three and nine (indicator set to two) becomes ATP days at the destination. This specifically allows the system to use the cheaper source for all orders from 10 days out and further, but allows the system to use the more expensive source between days three and nine to address any immediate shortages that may occur as a result of changes to forecast demand inside 10 days.

In situations where there are multiple secondary sources that might be considered as primary according to the Alternates as Primaries Indicator, only the secondary source with the earliest possible delivery day in its secondary release schedule is considered.

Note:

Alternates as Primaries functionality works for both warehouse and vendor sourcing. This means that both vendors and warehouses that are secondary sources may be used to address short term shortages at warehouse destinations even though they are not primary sources.

Delivery Day Demand Percent (DDP)

The DDP indicates what portion of a day's forecast demand can be met by the delivery on that day. The Derive Delivery-day Demand Percent flag indicates if the DDP should be calculated. If the answer is yes, then the DDP will be determined from a combination of on-shelf time and sales profile. If the answer is no, then the DDP will be the one specified directly by the user.

Deriving DDP from On-shelf Time and Sales Profile

When the DDP is to be derived, it is dependent upon both the On-shelf Time and Sales Profile. If either piece of data cannot be determined for a given SKU/Store and day, then the DDP should automatically be considered to be 100%.

On-shelf Times

Valid On-shelf times are those in the 24 hour clock format between 0001 hours and 2400 hours.

Sales Profile

The Sales Profile is provided by an external process which makes data available to AIP. The data provides the cumulative sales profile by hour for a given Sub-class/Location and day of week. The value stored represents what percentage of the day's sales has been completed at the specified hour. Valid values are required for hours numbered 1 to 24 inclusive. Valid percentage values are decimals between zero and 100 inclusive.

The percentage of sales should always be assumed to be zero for the zero hour. The percentage of sales for the completion of the 24th hour should always be 100 percent.

DDP Determination

Shelf Times can be expressed down to the minute. However, the Sales Profile is by hour. Shelf times that don't fall exactly on the hour need to derive the sales for the specified time by interpolation. The approach assumes a linear rate of sale between one specified hour and the next.

The Sales Percent for the on-shelf time is simply the Sales Percent associated with the specified On-Shelf hour plus a number of 60ths of the difference in sales percent between the next hour and the specified on-shelf hour. The number of 60ths is determined by the specified minutes.

Since sales percent is a percent complete the DDP is simply 1 - the sales percent.

Review Time (RT)

Those days at a location on which an order could be calculated are referred to as ATP days. The period of time which is the sole responsibility of a specific ATP day form a period of time called the Review Time (RT). The RT is determined by its start point and end point, referred to as the Beginning of Review Time (BRT) and End of Review (ERT) respectively. The RT contains all days (whole, partial or a combination of both) between the BRT and ERT.

DDP determines the BRT and ERT of the review time. The DDP indicates what portion of a day's forecast demand can be met by the delivery on that day. By implication a portion of that day determined as (1-DDP) must elapse before the delivery is made. If the DDP is for a day is 60%, then the BRT falls 40% (1-.6) through the day.

The BRT for all ATP days can be determined this way.

The ERT for a given ATP day is simply determined as the same point in time as the BRT of the very next ATP day.

If the next ATP day cannot be found due to a discontinuation date or a warehouse stop receiving date then the earliest date found defines the ERT.

Forecasted Demand over Review Time

A necessary pre-requisite for performing replenishment is a Forecast Demand stream. This stream is made available to replenishment for the lowest tier of the network (stores) only. It is necessary for Replenishment itself to determine demand on warehouses before they can be replenished in this process.

The output from replenishment of any SKU Pack/destination in the supply chain is a receipt plan showing desired order quantities to be delivered into the destination. The receipt plan is specific to a source and delivery date.

Each desired order quantity in the receipt plan also has a corresponding ship (or order) date. The ship date is calculated as the delivery date of the order quantity minus the lead time. Time shifting all orders to their ship date and summing them at the SKU Pack/source level yields a total of all orders that need to be shipped from the source. This total, for each SKU Pack/warehouse, represents the forecast demand stream for that SKU Pack on the warehouse.

In the case of a SKU which is capped, this demand still represents the shipments to meet the Unconstrained Receipt Plan at the destination store, without accounting for any capping constraints. This is because the warehouse should still order the entire amount needed from the stores. Thus, even if the stores cannot store all the stock they are potentially going to sell, the warehouses can meet that demand through direct shipments to customers.

Where order quantities for lower level supply chain tiers are generated outside of AIP the summed and time shifted quantities can be fed to the AIP warehouse as External Demand. This demand is added any demand generated from destinations within AIP. The warehouse, and any tiers above the warehouse, can then generate a receipt plan for the full warehouse demand.

Customer Orders over Review Time

Customer orders are orders placed by customers at a store for a specific quantity of a SKU for collection on a specific date. Customer Orders are entered in an external system (to AIP) and reach AIP as a desired quantity of a given SKU on a specific date, the date representing when the product is required in store. The acceptance of a customer order at a store is usually accompanied with a binding agreement that commits the store to meeting the order—over and above any other forecast of demand. This means that customer orders are treated as additional demand—on top of normal forecasted demand.

Back orders (BO) are another form of customer order. The customer's order is typically managed by Online Order Capture (OOC) and Order Management Systems (OMS). These systems mange the pickup or shipment method and associate the order to the appropriate location for fulfillment. An OOC system operates independently of a brick-and-mortar store but can leverage inventory available in stores and warehouses. There are times when OOC accepts an order believing there is inventory available. When the OMS releases the order for fulfillment and finds the inventory is not available it can manage the order as a back order. A back order has no committed fulfillment date and remains outstanding for fulfillment as soon as possible until it is fulfilled or the customer cancels.

CORT describes all Customer Orders and Back Orders to be considered by a single ATP day. In this instance, Review Time (by its strict definition) is not the period used to sum Customer Orders. Instead, they are summed for a period of time that starts on the ATP day in question and ends (but includes) the day before the next ATP day. Customer Orders are assumed to be zero (0) for warehouses.

Expected Receipts

Expected Receipts are those orders (Purchase Orders or Transfers) that have already been executed and therefore need to be considered as expected supply when planning future orders. Expected Receipts contain quantities of a specific SKU Pack due to arrive at a destination on a specified day.

Expected Receipts consist of the sum of two types of orders quantities:

On Orders quantity represents the quantity which has been executed, but as of yet there is no information regarding the physical shipment to the destination.
In Transits quantity represents the quantity which has physically shipped to the destination.

Product Life

Product Life (PL) refers to a number of days of life a product has upon receipt at a store. Once a product expires after a specified number of days, it can no longer be used to meet sales. The day of receipt at store is considered one day of a product's life so a product expires at the end of (the day of receipt + PL - 1 day).

Product Life is only relevant for a SKU when the following conditions are met:

The Store Use Inventory Aging Flag is True.
The Product Life is greater than zero (0).
The Product Life is less than or equal to the Store Inventory Aging Limit.

Product Life is never relevant at the warehouse.

Expected Spoilage

Expected Spoilage (ES) refers to a quantity of a SKU at a store whose product life expires and therefore becomes un-sellable at the end of a day. Expired quantities cannot be used for any sales after the day they expire.

Expected Spoilage is calculated for Current inventory and Planned or Expected receipts (PR and ER respectively) when product life is relevant and the Expected Spoilage Calculation Indicator is set to True and product life is relevant. All inventory arriving at a store as part of PR or ER is assumed to have full product life. Any remaining part of the receipt not used to fulfill demand spoils on (Day of receipt + Product Life - 1).

Inventory Capping and Store Available Space

The primary drivers of Inventory Capping functionality are as follows:

Cost of exceeding store capacity is significantly higher than any potential missed sales because of no stock at store. Thus the plan should never exceed the capacity constraint in terms of the inventory that a store can carry on any given day.
Store sources (warehouses) should still generate orders (if required) as if there was no capping at the store at all.

When managing to inventory caps, space is at a premium and overstocks within the store are costly and labor intensive to manage. As such, when replenishing, a receipt plan should be generated keeping in mind that it should not violate the capacity constraints at the store for any day. This is achieved by calculating an Available Space quantity for any given ATP day. The AIP processing logic ensures that the URP is never greater than the Available Space. Available Space is calculated (and hence Inventory Capping functionality is applied) for all SKUs that have the Capping Flag set to Yes. This functionality is applied at a SKU/Store/Day level.

Because the store space is limited, to ensure that shipments do not go over the SKU cap at a store, it is assumed that the forecasted sales are not realized over the execution window. The execution window represents the point in time when the order would have to be executed upon. This day is the Order Day, which is the lead time number of days earlier than the ATP day. The lead time in question is the lead time on the ATP day.

Also, any expected receipts over the planning horizon starting from the ATP day onwards, since they have already been committed, would be considered as taking up space as of the ATP day. In other words, space is reserved for them; hence we cannot use that space to plan for extra receipts on the ATP day.

Store Available Space Calculation

The Store Available Space calculation has the following characteristics:

Calculated only on ATP days
Maximum value cannot exceed the SKU Cap for the ATP day
Store Available Space is the SKU Cap on the ATP day minus
- Projected Inventory on the Order Day (day which is lead time (on the ATP day) days prior to the ATP day)
- Sum of Expected Receipts from the Order Day across the planning horizon
- Sum of Planned Receipts (if any) from the Order Day till the ATP day
If store available space is negative, it is set to zero (0).

Note:

SAS is first calculated during the replenishment (prior to capping) run and is updated as the PI and PR are updated during processing. The SAS calculated during capped replenishment still represents the most constrained available space, since the PR (and hence PI) can only be constrained further during reconciliation.

Since valid ATP days include only those days which have a shipment day of Today or in the future, t-LT(t) for an ATP day would always be a day in the planning horizon.

Allocation Boundaries

The basis for replenishment is the calculation of allocation boundaries. These boundaries represent varying levels of inventory required to satisfy particular requirements. The calculation of these boundaries varies significantly across the replenishment methods. The details of each boundary are described later according to replenishment method. The allocation boundaries are as follows:

Minimum Sales Stock

The Minimum Sales Stock (MSS) represents the minimum stock level that a SKU requires at a destination to satisfy forecast demand and customer orders, and achieve a projected inventory of zero (0) at the end of the review time period.
Safety Stock

Safety Stock (SS) is the amount of stock that should be kept on hand in an effort to mitigate the risk of out of stocks due to forecast variability.
Receipt Point

The Receipt Point (RP) represents the inventory level below which an order should be triggered.
Receive Up to Level

The Receive Up to Level (RUTL) denotes the target level of inventory when generating an order.
Maximum Sellable Quantity

The Maximum Sellable Quantity (MSQ) represents the level of inventory that can be sold within the life of a product.

Allocation Boundaries Calculation Variables

The following sections describe allocation boundaries calculation variables.

Safety Stock Minimums and Maximums

For replenishment methods that do not calculate Safety Stock based upon Sales, Safety Stock Minimums and Maximums provide a means of applying boundaries to limit the minimum and maximum values for calculated Safety Stock (SS). Before the boundaries can actually be applied to the calculated SS, the upper and lower boundaries themselves need to be determined. The calculation of the minimums and maximums do not vary by replenishment method.

The user may specify upper and lower limits, both in terms of numbers of days and numbers of units. Given that the upper and lower limits may be applied in both units and numbers of days at the same time, the following rules are used to determine which takes priority.

The lower limit to be applied to SS is simply calculated as the larger of either of the two specified Min values in units. That is the larger of either the Min Units or the Min Days (converted into units).
The upper limit to be applied to SS is simply calculated as the larger of either of the two specified Max values in units. That is the larger of either the Max Units or the Max Days (converted into units).
When neither the MinUnits nor Min Days measure has been specified by the user, the lower bound to be applied to SS is zero (0).
When neither the MaxUnits nor Max Days measure has been specified by the user, there is no upper boundary to be applied to SS.
Where the calculated lower boundary for SS exceeds the calculated upper boundary for SS, the lower boundary should be set to the upper boundary.

It is necessary to convert the Min and Max Day values into units so that they can be directly compared to the user specified Min and Max Unit values. Both the min and max day values, when specified, can be converted into units by summing the specified number of days of forecast demand commencing (and including) the Beginning of Review Time (BRT).

Supplier Compliance Safety Stock

The purpose of Supplier Compliance Safety Stock Uplift (SCSSU) is to address the poor record of a supplier for meeting their delivery obligations. The calculation of SCSSU involves summing demand for a period of time specified by the user as a number of days, Supplier Compliance Safety Stock Days (SCSSD). The beginning of the period over which the summation should take place is calculated as the current ATP day plus an additional number of days. This additional number of days is the maximum of the number of days in the review time for the current ATP day or the user-specified Inventory Selling Days (ISD) parameter.

Boundary Stock

Boundary Stock (BS) is a means of including other store promotion and presentation factors in addition to Safety Stock (SS) when considering the stock level at which a store is generated and how much stock to carry over and above demand. There is no presentation or promotional factors to consider for warehouse destinations. Therefore for warehouses, Boundary Stock is set to the Safety Stock calculated in a later section.

Warehouse Additional Stock

For warehouses, there are two additional methods of carrying inventory:

Baseline stock
Contingency stock

Baseline Stock represents additional inventory that should be carried for a discreet purposes and as such, it should be a quantity that is required over and above the system-calculated Safety Stock. As such, it is added onto the calculated Safety Stock. Given that the user can specify multiple reasons for adding baseline stock, the various baseline stock entries are added together before being added to the system calculated Safety Stock.

Contingency stock is a quantity that represents a minimum level of stock to be carried for contingency purposes. The user may enter multiple contingency stocks, but given that the contingency carried for one purpose can be used for another, the system uses the largest of the contingency stock entries and disregards the rest.

Just as the system chooses the largest of the contingency stock entries because stock carried for one purpose can be used for any un-forecasted event, the system applies this approach by comparing the largest of the user-specified contingency stocks against the system calculated Safety Stock. Safety Stock is then reset to the larger of the two.

Replenishment Methods Used in AIP

There are a number of replenishment methods that may be used to generate a replenishment plan in AIP. The different methods are appropriate for different kinds of situations.

The replenishment method that is used for a particular SKU or set of SKUs is typically chosen by super users who set the replenishment parameters or by a replenishment optimization system that suggests the most appropriate method and parameters. This method is used throughout the supply chain management process. The use of the replenishment method is monitored for effectiveness, and it is modified as needed over time.

The following replenishment methods can be used for store replenishment:

Min/Max
Time Supply
Dynamic
Hybrid
No Replenishment
No Safety Stock
Poisson Replenishment
Loaded SS Dynamic
Sales Week Range
Factor AWS

The following replenishment methods can be used for warehouse replenishment:

Min/Max
Time Supply
Dynamic
Hybrid
No Replenishment
No Safety Stock
Loaded SS Dynamic
Sales Week Range
Factor AWS
Factor ARS

The following sections discuss these replenishment methods, which includes an explanation of the allocation boundaries calculations.

Min/Max Replenishment Method

Min/Max is a simple, non-forecast-based replenishment method. It is generally used in the following types of situations:

When it is impossible to generate a reasonable forecast, such as a completely new type of product that cannot be modeled after anything else
For extremely slow-selling items where the minimum presentation levels constantly exceed the weekly demand

Safety Stock

Safety Stock in the Min/Max method consists of the minimum supply in units (MinS) multiplied by an increment percentage factor (Inc%).

Minimum stock level (user-defined parameter) is the minimum number of units required to have on hand to satisfy demand.

Increment Percent (user-defined parameter) is a multiplier for minimum and maximum stock that is designed to handle temporary fluctuation in demand for items on Min/Max replenishment without altering the established MIN and MAX values. Supplier Compliance Safety Stock is added into the safety stock. If the destination is a warehouse then once SCSSU has been added, the resulting Safety Stock value should be compared to the user-specified Contingency Safety Stock (in units). The user-specified Baseline Stock value (in units) should then be added to the larger of the two safety stock values.

Receipt Point

Receipt Point in the Min/Max method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

For stores, Demo Stock (DS) must be added to the previously-calculated RP. However, the value may not exceed the sum of MSQ plus Demo Stock.

Also, when the product/store is constrained by shelf capacity (Shelf Capacity Flag is True), the current RP value is compared to the sum of Shelf Capacity (SC), CORT and Demo Stock. RP is set to the lesser of the two.

Receive Up to Level

For the Min/Max replenishment method, Receive Up to Level (RUTL) is calculated by adding Customer Orders over Review Time (CORT) to the max of Boundary Stock (BS) and the product of the Maximum Stock (MaxS) and Increment Percent (Inc %) values.

For Stores, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Also, when the product/store is constrained by shelf capacity (Shelf Capacity Flag is True), the current RUTL value is compared to the sum of Shelf Capacity (SC), CORT and Demo Stock. RUTL is set to the lesser of the two.

Maximum Sellable Quantity

If the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

If the Freshness Flag is set to False, the Customer Orders are summed for the life of the product, then added to the sum of Forecast Sales over the life of the product and Back Orders to determine the final MSQ.

MSQ is dependent upon Product Life (PL) and the MSQ Constraint Indicator. Where Product Life is deemed to be not relevant for a SKU or the MSQ Constraint Indicator is False, or both, then the MSQ is considered infinite (Unbounded).

Note:

Maximum Sellable Quantity (MSQ) only applies to stores. MSQ for warehouse is unbounded because Product Life (PL) is not relevant at warehouses.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Min/Max replenishment method is calculated as the minimum between MSQ and the sum of Safety Stock and Customer Orders over the Review Time.

Finally when the product/store is constrained by shelf capacity (Shelf Capacity Flag is True) the calculated MSS for stores may not exceed the level required to fill the shelf to capacity and meet CORT. Where it does, MSS is set to Shelf Capacity (SC) + CORT.

Time Supply Replenishment Method

The Time Supply replenishment method allows the user to maintain a minimum and maximum amount of stock in terms of days of supply.

The Time Supply Horizon Parameter

The Time Supply Horizon parameter is an optional parameter used in Safety Stock and receives up to level calculations for this method. A projected daily average rate of sale is calculated based on the forecasted demand over the time supply horizon. This parameter can be used to smooth spiky forecasts.

If the Time Supply Horizon parameter is undefined, the total of the forecast over the minimum and maximum time supply days is used instead. This results in a truer calculation of the demand since the forecasted demand over the minimum supply days and maximum supply days may have varying rates of sale.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Time Supply replenishment method is calculated as the minimum between MSQ and the sum of Safety Stock and Customer Orders over the Review Time.

When the product/store is constrained by shelf capacity (Shelf Capacity Flag is True) the calculated MSS for stores may not exceed the level required to fill the shelf to capacity and meet CORT. Where it does, MSS is set to Shelf Capacity (SC) + CORT.

Safety Stock

Safety Stock (SS) in the Time Supply method consists of calculating a stock level based on the forecasted demand or an average rate of sales.

If a Time Supply Horizon (TSH) has been specified, then this method establishes an Average Rate of Sale (ROS) over the time supply horizon period starting (and including) the beginning of review time. This average is then multiplied by the Minimum Time Supply Days (MinTS).

If a time supply horizon is not specified, then the method sums the forecasted demand over the Minimum Time Supply Days (MinTS) starting with (and including) the beginning of the review time.

The result of either of the preceding calculations is then checked to ensure that it falls between the SSMIN and SSMAX boundaries determined in the Safety Stock Minimums and Maximums section. Then, the user-defined Safety Stock level factor (SSLF) is added.

Receipt Point

Receipt Point in the Time Supply method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT)

For Stores, Demo Stock (DS) must be added to the calculated RP. However, the value may not exceed the sum of MSQ plus Demo Stock.

Receive Up to Level

Receive Up to Level (RUTL) in the Time Supply method consists of calculating a target stock level used in generating an order based on the forecasted demand or an average rate of sales.

If a time supply horizon is not specified, then the method sums the forecasted demand over the Maximum Time Supply Days (MaxTS) period starting (and including) the beginning of the review time.

The result of either of the preceding calculations is added to the Safety Stock level factor (SSLF). It is then compared against the Boundary Stock. The larger of the two values are selected and Customer Orders over Review Time (CORT) is added to get the RUTL value..

For Stores, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Maximum Sellable Quantity

If the Freshness Flag is set to False, the Customer Orders are summed for the life of the product, then plus the sum of Forecast Sales over the life of the product and Back Orders to determine the final MSQ.

Dynamic Replenishment Method

The goal of the Dynamic Replenishment method is to replenish only the quantity required to meet user specified customer service level. It is a statistical method that assumes the forecast demand observes normal distribution. The dynamic method minimizes the amount of Safety Stock on hand while meeting specified service levels. Higher service levels results in higher levels of Safety Stock. Forecast accuracy is also important for this method. Products with high standard deviations require more stock to cover the deviations.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Dynamic Replenishment method is calculated as the minimum between MSQ and the sum of Forecast Demand and Customer Orders over the Review Time.

Calculated MSS may not exceed Maximum Sellable Quantity (MSQ). Where it does, MSS is set to MSQ.

Safety Stock

Dynamic Safety Stock (SS) is a statistical method that assumes the forecast demand observes normal distribution. For this method, it is first necessary to establish a value for Standard Deviation over Review Time. The calculation of Standard Deviation over Review Time depends upon which of the Daily or Weekly Store Standard Deviation external feeds are provided. In both cases, the calculated Standard Deviation makes use of a Forecast Demand Index value.

To determine Standard Deviation over Review Time, it is first necessary to establish the Standard Deviation of the Daily Forecast Error at the warehouse level. The Daily Forecast Error is the difference between the Forecast Demand on the warehouse and the Actual Demand for a given day. Forecast Demand on the warehouse for the specified day is the summation of all destinations' orders time-shifted to the warehouse ship date. The Actual Demand on the warehouse for the specified day is the total volume of all destinations' orders released to the warehouse on the specified day.

The timing of the capture of the Forecast Demand data is crucial. The Forecast Demand for a specific day should be captured on release of the ATP day whose review time includes that day. It is possible because of changes in lead times for the Forecast Demand for a given day to be captured multiple times. In this instance and other instances where there could be ambiguity over the capture of data—the last captured value is the one that is used.

When determining the Standard Deviation of the DailyErrorWH term, the DailyErrorWH is assessed over a user-determined historic number of days called the Average Daily Demand Period (ADDP). The Average Daily Demand (ADD) can then be determined for the ADDP using the Actual Demand (AD) the capture of which is described in the previously.

Where the ADDP is at least seven days, the Standard Deviation of the Daily Forecast Error is calculated the Standard Deviation of the DailyErrorWH term using a standardized formula. Where the ADDP is less than seven days, the Standard Deviation of the Daily Forecast Error is simply calculated as a function of the ADD over the ADDP.

The calculation of the Standard Deviation over Review Time makes use of a Forecast Demand Index to address seasonality issues. The calculation of FDI is based upon the determination of the Forecast Demand over Review Time as a function of averaged demand over a similar number of days. The Average Daily Demand (ADD) is the same value daily value that was assessed in earlier calculations. The calculation of FDI is based upon the setting of a calculation indicator. If the indicator is True, then FDI is calculated. Otherwise, the FDI value used in the calculation can be loaded. The default value of the loaded measure is 1. This maintains the integrity of calculations when no value is loaded.

The use of a Forecast Demand Index (FDI) addresses seasonality issues. The calculation of FDI is based upon the determination of the Forecast Demand over Review Time as a function of averaged demand over a similar number of days. The calculation of FDI is based upon the setting of a calculation indicator. If the indicator is True, then FDI is calculated. Otherwise, the FDI value used in the calculation can be loaded. The default value of the loaded measure is 1. This maintains the integrity of calculations when no value is loaded.

Once the Standard Deviation Over Review Time has been determined, an Acceptable Unit Loss for the Review Time (AULRT) should be calculated. The AULRT represents the quantity of stock outs for a given product that it is acceptable to incur based on the specified service level over the review time.

A standard lookup function using a normal distribution table is then used to determine a factor which is multiplied by the standard deviation for the review time. The acceptable unit loss (AUL) over the review time is used to find the Z-factor needed for the Statistical Safety Stock Calculation. The result of the lookup is multiplied by the standard deviation over the review time to determine the statistical Safety Stock.

The result is then checked to ensure that it falls between the SSMIN and SSMAX boundaries determined in the Safety Stock range section. Finally, the Safety Stock level factor (SSLF) is added and the result is checked to ensure that it does not drop below zero (0).

Finally, Supplier Compliance Safety Stock Uplift is added to the calculated Safety Stock.

If the destination is a warehouse then once SCSSU has been added, the resulting Safety Stock value should be compared to the user-specified Contingency Safety Stock (in units). The user-specified Baseline Stock value (in units) should then be added to the larger of the two safety stock values.

Receipt Point

Receipt Point in the Dynamic method is calculated in one of two ways based on the Inventory Boundaries as Aesthetic Min value.

When Store Inventory Boundaries as Aesthetic Min is set to False, the RP is calculated as the sum of demand over the review time, plus the previously calculated Boundary Stock and CORT.

When Store Inventory Boundaries as Aesthetic Min is set to True, the RP is calculated as the max of demand over the review time plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). CORT is added to the result.

For Stores, add Demo Stock and check the value does not exceed plus Demo Stock. Finally, when the product/store is constrained by shelf capacity (Shelf Capacity Flag is True), the current RUTL value is compared to the sum of Shelf Capacity (SC), CORT and Demo Stock. RUTL is set to the lesser of the two.

When Shelf Capacity is not relevant for the SKU:

Receive Up to Level

For Receive Up to Level (RUTL) using the Dynamic Replenishment method, demand is assessed over a number of days. This number of days is the larger of either the number of days in the review time, or the user-specified Inventory Selling Days (ISD).

The RUTL is then calculated in one of two ways based on the Inventory Boundaries as Aesthetic Min value.

When Store Inventory Boundaries as Aesthetic Min is set to False:

The RUTL is then calculated as the sum of demand over the calculated number of days commencing the beginning of the review time plus the previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True:

The RUTL is calculated as the max of demand over the calculated number of days plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

For Stores, Customer Orders (CORT) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

For Stores, add CORT and Demo Stock and check the value does not exceed MSQ, constraining by shelf capacity when relevant.

Maximum Sellable Quantity

For the Dynamic Replenishment method, if the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

If the Freshness Flag is set to False, the Customer Orders are summed for the life of the product, then add the sum of Forecast Sales over the life of the product and Back Orders to determine the final MSQ.

Hybrid Replenishment Method

The Hybrid replenishment method is a combination of Dynamic and Time Supply replenishment methods. The main difference between the Hybrid and Dynamic methods is the calculation of Safety Stock. In the Hybrid method, the Safety Stock is calculated using the same algorithm used in the Time Supply method.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Hybrid replenishment method is calculated the minimum between MSQ and the sum of Forecast Demand Over review time and Customer Orders over the Review Time.

Safety Stock

Safety Stock (SS) in the Hybrid method consists of calculating a stock level based on the forecasted demand or an average rate of sales.

If a time supply horizon is not specified, then the method sums the forecasted demand over the Minimum Time Supply Days (MinTS) starting with (and including) the beginning of the review time.

The result of either of the preceding calculations is then checked to ensure that it falls between the SSMIN and SSMAX boundaries determined in the Safety Stock Minimums and Maximums section. Finally, the user-defined Safety Stock Level Factor (SSLF) is added.

Receipt Point

Receipt Point in the Hybrid method is calculated in one of two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False:

The RP is then calculated as the sum of demand over the review time, plus the previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True:

The RP is then calculated as the max of demand over the review time plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

Add Demo Stock and check the value does not exceed MSQ, plus Demo Stock.

Receive Up to Level

For Receive Up to Level (RUTL) using the Hybrid replenishment method, demand is assessed over a number of days. This number of days is the larger of either the number of days in the review time, or the user-specified inventory selling days (ISD).

The RUTL is then calculated in one of two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False

The RUTL is calculated as the sum of demand over the calculated number of days commencing the beginning of the review time plus previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True

The RUTL is then calculated as the max of demand over the calculated number of days plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

For Stores, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Maximum Sellable Quantity

For the Hybrid replenishment method, if the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

No Safety Stock Method

The No Safety Stock Method allows retailers to order forecast only, no safety stock. This method will be leveraged for planning non-stockholding stores, stockless warehouses and any business cases where safety stock is not desired.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the No Safety Stock Method is calculated as the minimum between MSQ and the sum of Forecasted Demand Over review time and Customer Orders over the Review Time.

Safety Stock

For No Safety Stock method, no safety stock is ordered. Set safety stock to zero.

Receipt Point

Receipt Point in the No Safety Stock method is calculated in one of the two ways based on the Inventory Boundaries as Aesthetic Min value.

When Store Inventory Boundaries as Aesthetic Min is set to False, the RP is then calculated as the sum of demand over the review time, plus the previously calculated Boundary Stock and CORT.

When Store Inventory Boundaries as Aesthetic Min is set to True, the RP is then calculated as the max of demand over the review time plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

Add Demo Stock and check the value does not exceed MSQ, plus Demo Stock.

Receive Up to Level

For Receive Up to Level (RUTL) using the No Safety Stock replenishment method, demand is assessed over a number of days. This number of days is the larger of either the number of days in the review time, or the user-specified inventory selling days (ISD).

The RUTL is then calculated in one of the two ways based on the Inventory Boundaries as Aesthetic Min value.

When Store Inventory Boundaries as Aesthetic Min is set to False, the RUTL is then calculated as the sum of demand over the calculated number of days commencing the beginning of the review time plus the previously calculated Boundary Stock and CORT.

When Store Inventory Boundaries as Aesthetic Min is set to True, the RUTL is then calculated as the max of demand over the calculated number of days plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

For Stores, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Maximum Sellable Quantity

For the No Safety Stock replenishment method, the Maximum Sellable Quantity calculated in one of the following two ways. If the Freshness Flag is set to true, then Maximum is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time. If the Freshness Flag is set to false, the Customer Orders are summed for the life of the product, and then plus the sum of Forecast Sales over the life of the product and Back Orders to determine the final MSQ.

No Replenishment Method

The No Replenishment method is provided for cases when replenishment calculations need to be turned off for a specific SKU/location/day.

This method is used for periods when a location does not want to have any replenishment performed.

Where No Replenishment is used, all allocation boundaries (MSS, SS, RP, RUTL, AND MSQ) are set to zero (0).

Poisson Replenishment Method

Note:

The Poisson replenishment method is for stores only.

The Poisson replenishment method is identical to the Dynamic Replenishment method, except for the way the Safety Stock is derived. Poisson uses a Poisson distribution lookup table fed by the demand over the review time and the desired service level as parameters. Poisson does not incorporate forecast error.

Required Data

The following data must be loaded in order to use the Poisson replenishment method. Please refer to the Oracle Retail Advanced Inventory Planning Implementation Guide for details on loading data.

Interval Hierarchy
Poisson Distribution Table

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Poisson replenishment method is calculated as the minimum between MSQ and the sum of Forecasted Demand Over review time and Customer Orders over the Review Time.

Safety Stock

Poisson is a statistical method that uses a Poisson distribution lookup table. The lookup function is fed the demand over the review time and the desired service level as parameters. The Poisson table is loaded from an external source.

The result is then checked to ensure that it falls between the SSMIN and SSMAX boundaries determined in the Safety Stock range section. Finally, the Safety Stock Level Factor (SSLF) is added and the result is checked to ensure that it does not drop below zero (0).

Finally, Supplier Compliance Safety Stock Uplift is added to the calculated Safety Stock.

Receipt Point

Receipt Point in the Poisson method is calculatedin one of the two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False

The RP is then calculated as the sum of demand over the review time, plus the previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True

The RP is then calculated as the max of demand over the review time plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

Add Demo Stock and check the value does not exceed MSQ, plus Demo Stock.

Finally, when the product/store is constrained by shelf capacity (Shelf Capacity Flag is True), the current RUTL value is compared to the sum of Shelf Capacity (SC), CORT and Demo Stock. RUTL is set to the lesser of the two.

Receive Up to Level

For Receive Up to Level (RUTL) using the Poisson replenishment method, demand is assessed over a number of days. This number of days is the larger of either the number of days in the review time, or the user-specified inventory selling days (ISD).

The RUTL is then calculated in one of the two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False

The RUTL is then calculated as the sum of demand over the calculated number of days commencing the beginning of the review time plus the previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True

The RUTL is then calculated as the max of demand over the calculated number of days plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

Next, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Add CORT and Demo Stock and check the value does not exceed MSQ, constraining by shelf capacity when relevant.

Maximum Sellable Quantity

For the Poisson replenishment method, if the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

Loaded SS Dynamic Replenishment Method

The Loaded SS Dynamic Replenishment method allows the user to provide a custom Safety Stock quantity from an external system.

Required Data

The following data must be loaded in order to use the Loaded SS Dynamic Replenishment method. Please refer to the Oracle Retail Advanced Inventory Planning Implementation Guide for details on loading data.

Store Loaded Safety Stock

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Loaded SS Dynamic Replenishment method is calculated as the minimum between MSQ and the sum of Forecasted Demand Over review time and Customer Orders over the Review Time.

Calculated MSS may not exceed Maximum Sellable Quantity (MSQ). Where it does, MSS is set to MSQ.

Safety Stock

Safety Stock in the Loaded SS Dynamic method does not calculate a Safety Stock but instead loads it from an external system. As a result, no attempt is made to confine the loaded value to the Safety Stock minimum and maximum limits. Supplier Compliance Safety Stock Uplift is added to the loaded Safety Stock.

Receipt Point

Receipt Point in the Loaded SS Dynamic method is calculated in one of the two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False

The RP is then calculated as the sum of demand over the review time, plus the previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True

The RP is then calculated as the max of demand over the review time plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

For stores, add Demo Stock and check the value does not exceed plus Demo Stock.

Receive Up to Level

For Receive Up to Level (RUTL) using the Loaded SS Dynamic Replenishment method, demand is assessed over a number of days. This number of days is the larger of either the number of days in the review time, or the user-specified inventory selling days (ISD).

The RUTL is then calculated in one of the two ways based on the Inventory Boundaries as Aesthetic Min value:

When Store Inventory Boundaries as Aesthetic Min is set to False

The RUTL is then calculated as the sum of demand over the calculated number of days commencing the beginning of the review time and then adding previously calculated Boundary Stock and CORT.
When Store Inventory Boundaries as Aesthetic Min is set to True

The RUTL is then calculated as the max of demand over the calculated number of days plus Safety Stock (SS), or Presentation Stock (PS) plus Promotional Presentation Stock (PPS). Then add CORT to the result.

For Stores, Demo Stock (DS) must be added to the previously-calculated RUTL. However, the value may not exceed the sum of MSQ plus Demo Stock.

Maximum Sellable Quantity

Sales Week Range Replenishment Method

The Sales Week Range replenishment method allows the user to define the Safety Stock (SS), Receipt Point (RP) and Receive Up to Level (RUTL) based on sales history. The basic calculation generates Safety Stock based on the sum of sales over a user defined period in history.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Sales Week Range replenishment method is calculated as the minimum between MSQ and the sum of Safety Stock and Customer Orders over the Review Time.

Calculated MSS may not exceed Maximum Sellable Quantity (MSQ). Where it does, MSS is set to MSQ.

Safety Stock

Safety Stock for the Sales Week Range replenishment method is calculated as the total sales in a historical period specified by the user. The historical period is defined by a start and end date specified by the user.

Table 11-1 shows the two options for the Sales Week Range method controlled by the Roll Weeks Flag.

Table 11-1 Historical Period for the Roll Weeks Flag

If the Roll Weeks Flag is Set to...	Then the Historical Period...
False	Is used in the summation of sales is static, where the sales are summed over the date range specified for each day the Safety Stock is calculated.
True	Moves forward a week for each new week encountered in the planning horizon. The sales period does not roll forward beyond first day in the planning horizon.

The Roll Weeks Flag is time-phased, and so the value used through the entire planning horizon for the current run of Replenishment is the value valid for today (the current day).

The Safety Stock is calculated as the sum of historic sales between the current Start and End dates. The result is then checked to ensure that it falls between the optional sales min (SLSMIN) and sales max (SLSMAX) specified by the user (Store Sales Min, Store Sales Max, Warehouse Minimum, and Warehouse Maximum).

Finally, Supplier Compliance Safety Stock Uplift is added to the calculated Safety Stock.

Note:

If the Store Use Like SKU is set to True, the historical daily sales of an alternative Like SKU may be used in lieu of the sales of the SKU being replenished. If the use of a Like SKU is permitted and has been specified, then its historical sales are used if the Store Like SKU End Date is greater than the first day in the planning horizon.

Receipt Point

Receipt Point in the Sales Week Range method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

For Stores, addDemo Stock and check the value does not exceed MSQ, plus Demo Stock.

Receive Up to Level

Receive Up to Level in the Sales Week Range method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

For Stores, add Demo Stock and check the value does not exceed MSQ, plus Demo Stock.

Maximum Sellable Quantity

For the Sales Week Range replenishment method, if the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

Factor AWS Replenishment Method

The Factor AWS replenishment method allows the user to define the Safety Stock (SS), Receipt Point (RP) and Receive Up to Level (RUTL) based on average weekly sales. Sales are averaged over a user designated time period to arrive at an Average Weekly Sales (AWS) value. The AWS is multiplied by a user designated factor to generate Safety Stock.

Minimum Sales Stock

The Minimum Sales Stock (MSS) for the Factor AWS replenishment method is calculated as the minimum between MSQ and the sum of Safety Stock and Customer Orders over the Review Time.

Calculated MSS may not exceed Maximum Sellable Quantity (MSQ). Where it does, MSS is set to MSQ.

Safety Stock

Safety Stock for the Factor AWS replenishment method calculates Average Weekly Sales (AWS) and multiplies it by a user specified factor. AWS is calculated using sales in a historical period specified by the user. The historical period is defined by a start and end date specified by the user.

Table 11-2 shows the two options for the Factor AWS method controlled by the Roll Weeks Flag.

Table 11-2 Historical Period for Roll Weeks Flag

If the Roll Weeks Flag is Set to...	Then the Historical Period...
False	Is used in the AWS calculation is static, where the AWS is calculated using the date range specified for each day the Safety Stock is calculated.
True	Moves forward a week for each new week encountered in the planning horizon. The sales period does not roll forward beyond first day in the planning horizon.

The Roll Weeks Flag is time-phased, so the value used through the entire planning horizon for the current run of Replenishment is the value valid for today (the current day).

The Safety Stock is calculated as the average of the historic daily sales between the current Start and End dates. This number is then multiplied by seven to achieve a weekly value. The weekly value is then multiplied by a user specified Average Weekly Sale Factor. The result is then checked to ensure that it falls between the optional Store Sales Min (SLSMIN) and Store Sales Max (SLSMAX) specified by the user.

Note:

Receipt Point

Receipt Point in the Factor AWS method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

For Stores, add CORT and Demo Stock and check the value does not exceed MSQ, plus Demo Stock.

Receive Up to Level

Receive Up to Level in the Factor AWS method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

For Stores, add Demo Stock and check the value does not exceed MSQ plus Demo Stock.

Maximum Sellable Quantity

For the Factor AWS replenishment method, if the Freshness Flag is set to True, then Maximum Sellable Quantity (MSQ) is calculated as the sum of forecasted demand over the product life starting from the beginning of the review time and Customer Orders over Review Time.

Factor ARS Replenishment Method

Note:

The Factor ARS Replenishment Method only applies to warehouses.

The Factor ARS method allows the user to define the Safety Stock based upon an Average Rate of Sale (ARS) loaded from an external system.

There are two variants of this method. One variant keeps the time period specified by the user static, regardless of which week in the planning horizon is being planned. The second variant rolls the time period specified by the user forward one week at a time as Replenishment progresses through the planning horizon. This keeps the time period specified by the user relative to the day being planned.

Along with the user specified Start and End dates there is a Roll Weeks Flag. This flag determines which variant of the method is being used. It is important to understand how the rolling of weeks occurs if the user is to fully comprehend the impact of setting this flag to True.

The Replenishment process actually steps through every day in the planning horizon. However, it only runs boundary calculations and considers ordering on ATP day. The process of stepping through every day in the planning horizon is significant because it is this action that enables the rolling of weeks if the Roll Weeks Flag is set to True.

If the Rolls Weeks Flag is set to True, then as Replenishment walks through each day in the planning horizon, it checks to see if the day is the first day in a week (Sunday). If it is the first day in the week and not the very first day in the planning horizon, the Start Date and End Date specified by the user are moved forward exactly one week (seven days is added to both dates) as long as both dates still remains in the past. If adding seven days places one or both dates on or after the today's date, neither date is moved.

If the Roll Weeks Flag is False, then the Start and End date specified by the user remain static throughout the Replenishment Process.

Required Data

The following data must be loaded in order to use the Factor ARS replenishment method. Please refer to the Oracle Retail Advanced Inventory Planning Implementation Guide for details about loading data.

Store Average Weekly Rate of Sale
Total Store Average Rate of Sales

Safety Stock

Safety stock in the Average Rate of Sale Method is calculated by adding the loaded ARS value to a user-specified independent ARS value. The resulting value is assumed to be weekly and is divided by 7 to achieve a daily number before being multiplied by a user-specified ARS factor.

Receipt Point

Receipt Point (RP) for the Average Rate of Sale method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

Receive Up to Level

The Receive Up to Level (RUTL) for the Average Rate of Sale method is set to the calculated Boundary Stock (BS) plus Customer Orders over Review Time (CORT).

Calculate an Unconstrained Receipt Plan

The following sections describe how to calculate an unconstrained receipt plan.

Current Inventory and Projected Inventory

Projected Inventory (PI) represents a projection of an available inventory position at a specific time for a location. PI is a beginning of day inventory position. However, with intra-day deliveries comes the need to assess an inventory position at any point in the day. Therefore, an additional type of PI position has to be considered—one that represents inventory available at Beginning of Review Time.

Today is a special case for PI, since there is no need to project an inventory position for today if a Current Inventory (CI) feed is available from an external system. However, the following must be considered when determining PI for today:

Transfers In the Well (TIW) must be deducted from CI because the inventory required to satisfy them as not already been removed from the CI and this inventory cannot be considered available to meet future demand.
Actual Sales (AS) for today must be added back into the inventory. Usually there will be no AS but in the case of an intra-day batch run, some inventory may already have been sold and so needs to be added to determine a revised Beginning of Day position for today.

Where a Current Inventory Feed is not available for today, then a contingency Projected Inventory calculation is used.

The contingency PI is calculated as yesterday's calculated PI for today plus the difference between forecast demand (sales) for yesterday and Actual Sales for yesterday. If actual sales are not available for yesterday, then PI for today should simply be set to yesterday's projection of PI for today.

Projected Inventory on day t is calculated as the PI of the previous day (t- 1) plus the net impact of all demand, supply and spoilage on the previous day. The sequence in which these elements are considered is important. Note that not all elements are applicable to both warehouses and stores. Where an element is not applicable to a given destination type, it is considered to be zero (0).

First, the demand that cannot be satisfied by any receipts must be subtracted from the starting position. This demand is Forecast Demand. This demand should be multiplied by (1-DDP) so that only the part of demand that cannot be satisfied by that day's receipts is decremented. The demand that happens before any receipt of inventory takes places can only be satisfied by the starting inventory for that day. Any demand that cannot be satisfied by the opening inventory is simply lost.

Receipts are considered next. All planned and expected receipts are added.

Once the receipts have been added, the remaining part of that day's demand is subtracted. The remaining part of that day's demand is determined by taking Forecast Demand and multiplying it by the DDP then add to it Customer Orders to get the sum of demand.

Unlike Forecast Demand, CO can rely upon any deliveries that day that may be required to satisfy them.

Allocations in the Well (AIW) should be deducted. Although AIW is considered demand, it is different from other elements of demand because it is usually the result of an allocation (created outside of AIP) and is handled differently by the warehouse. If there is insufficient opening inventory to satisfy that day's orders, the orders are unlikely to be picked until that day's receipts have arrived; therefore, the calculation reflects the deduction after that day's receipts have arrived.

Expected Spoilage is then deducted. Finally, any Residual Excess Quantity is then deducted. The entire calculation is then capped so it may not drop below zero (0).

The starting point for calculating the PI position at the Beginning of Review Time (BRT) on day t is the Beginning of Day PI position on day t. Therefore, if a BRT PI position needs to be calculated on day t, the first step is to determine the PI on day t using the logic described above.

The PI position at the BRT is then determined by subtracting, from the starting position, the demand that will take place prior to the arrival of any inventory on that day (t). This demand is calculated by multiplying the Forecast Demand on day t by (1-DDP) for day t.

Net Inventory

Net Inventory (NI) is a prediction of the total inventory available over a given period of time—usually the review time associated with a specific ATP day. All demand associated with any previous ATP day and its associated review time must have already been deducted so that the NI position represents the available inventory to meet any demand in the current review time.

All NI calculations are based upon a PI position at the beginning of the review time and then consider the impact of Expected Receipts (ER), Allocations in the Well (AIW), Holdback, Substitute Quantities, related SKU inventory (Event SKUs), and Expected Spoilage (ES) over the period from the current ATP day to the day before the next ATP day.

Expected Receipts

Expected receipts are those quantities already expected to arrive into the location. These are summed between the current ATP day and the day before the next ATP day.

Allocation in the Well

Allocations in the Well represent orders sitting in the warehouse system which are to be honored a future day. These orders consume warehouse inventory and they must therefore be deducted from the inventory position when determining how much inventory is available to meet future demand. These orders are satisfied after all inventory has arrived. They too are summed between the current ATP day and the day before the next ATP day.

Spoilage calculations

Expected Spoilage is calculated as an end of day quantity. It represents the inventory that will be discarded because it has not been sold within its life. ES is summed between the current ATP day and the day before the next ATP day.

Warehouse Holdback Quantity

You may specify a Holdback quantity representing a quantity of inventory that is put aside at the warehouse and considered not available to meet demand. The user specifies if Holdback should be included in the NI calculation.

Substitute Quantity

Substitute Quantities represent the additional demand expected as a consequence of other SKUs experiencing stock outs at the same location. The quantities are summed over the review time to further reduce the NI position.

Calculate Ideal Receipt Quantity

The Ideal Receipt Quantity (IRQ) calculation is the process of actually determining what the order quantity should be. This is the same process for all methods except No Replenishment. For No Replenishment, the IRQ is zero (0).

For Stores, the determination of IRQ is impacted by the presence of User Specified Allocations (USA). If the ATP day being planned is before the On Supply Date for the SKU/Store or is after the Off Supply Date but within 42 days after it, then the IRQ should be set as the total of all USAs within the review time for the ATP day.

When the ATP day being planned falls within the On and Off Supply dates, the Net Inventory (NI) is compared to the Receipt Point (RP). If the Net Inventory is less than the Receipt Point, then the initial IRQ is calculated as the Receive Up to Level (RUTL) minus the Net Inventory.

If there are USAs in the review time of the ATP being planned, then the store IRQ is further modified. This modification depends upon the type of the first USA encountered in the review time. If the type is replace, then the sum of all the USAs in the review time replaces the calculated IRQ. If the type is Add, then the sum of all the USAs in the review time is added to the calculated IRQ.

Finally, the IRQ is checked against Customer Orders over Review Time (CORT). If Product Life is relevant for the SKU and the Freshness Flag is True, then the IRQ must be no less than CORT. If IRQ is less than CORT, the IRQ is set to CORT. This guarantees the freshest product is available to meet CORT.

If product life is not relevant for the SKU or the freshness flag is not True, then the inventory position including IRQ (NI + IRQ) must be no less than CORT. If it is, then the IRQ is set as (CORT - NI). This ensures there is sufficient inventory once the order has been received to meet CORT.

Allocate Order to Source

The initial Ideal Receipt Quantity (IRQ) is the total calculated for a destination. It is necessary to allocate the quantity across the one or more sources who can deliver on the Available To Plan day. This activity is simple for stores which always have a single source for an ATP day. All of the order is allocated to that source. For warehouses, this activity involves allocating quantity to one or more sources. The allocation is performed based on the following two pieces information:

The Source Splits information mastered in the Data Management Application.

The Source Splits indicate how a warehouse destination's orders should ideally be split long-term across all the available sources that can deliver the product. The orders on any given day need not hit the target percentages allotted to each source; the aim is to achieve the target splits over time, and therefore allow for the likelihood that there are occasions when only one of the sources can deliver on a day, hence receiving one hundred percent of the order.
Historical Order History for each Source Split captured by the Order management Application.

The historical orders, maintained at a demand group level, represent the total unit orders for each source split given to each source prior to the current batch run.

Only sources that have a positive warehouse percentage split gets orders. When there is only one source with a positive percentage who can deliver on a given ATP day, that source gets all the orders. However, if there are several sources, then the IRQ needs to be divided amongst them in a way that gets the long term history of orders (including the order now being split) as close as possible to the target percentages.

When allotting order quantities to a given source, this is done in order multiple quantities. These quantities are always in units. The order multiple is specified differently for warehouses than stores, but the order multiple to be used is always the one associated with the preferred ordering pack-size specified by the user. For a store, the appropriate Order Multiple is the pack-size of the Store Format Pack-size associated with the SKU and store, or the pack-size of the Store Format Pack-size exception if a relevant one exists. For a warehouse, the appropriate Order Multiple is the user-specified Order Multiple associated with the Location Orderable Unit (LOU). Both the Order Multiple and the LOU are specified by the user in the Data Management module. The LOU is entered in cases, and so this value needs to be converted to eaches prior to its use.

It is important to understand that one or more of the sources allotted a percentage of the demand for a warehouse may not be available to deliver on an ATP day. Therefore, the split percentages of those sources who can deliver on the ATP day must be normalized so that they total one hundred percent. This ensures that the entire order is accounted for.

Replenishment then goes through a number of calculations to split the IRQ across the available sources in whole order multiple quantities. Each order multiple allotted to a source represents part of the order to be delivered into the destination from that source. Each time a whole order multiple is added to a Source's order, the IRQ of the destination is reduced by that same order multiple quantity. This process stops when all whole order multiples have been allotted to sources and allotting another order multiple to a source would reduce the IRQ of the destination below zero (0).

After this process is complete one or more of the available sources on the ATP day is allotted order quantities that represent one or more whole order multiples. Unless the original IRQ happened to be a multiple of the sources order multiples, then there is a remaining part of the IRQ that is smaller than one of the sources' order multiples. A rounding decision now needs to be made about this remaining part of the IRQ.

Round Ideal Receipt Quantity

Once all whole order multiples have been allotted to sources, a rounding decision must be made concerning the remaining IRQ of the destination. This rounding decision uses the order multiple, which can vary by source, making it necessary to identify which source is the one that may receive an additional order multiple if one is needed.

If there is only one source on the ATP day, then that source is the one that receives an additional order multiple should the decision be made to round up. However, if there are multiple sources on the ATP day, then the source that is farthest away from its target split percentage is the one selected.

If the decision is made to round down, then the remaining quantity is discarded and no more order multiples are added to any source's existing order.

Regardless of which source is selected, the rounding decision is based upon a number of criteria, any of which could cause another order multiple to be ordered. Once it is determined that another order multiple is required, another order multiple is allotted to the chosen source and all further checking to see if another order multiple is required is bypassed.

Store Order Rounding

For a store, there is only ever a single source on an ATP day. Therefore, at this stage in the process, the IRQ allocated to a source already contains any whole order-multiples for that source. Should it be determined that another order multiple is required, it will be added to the IRQ of the source.

The order rounding for stores checks a number of conditions, any of which could cause another order-multiple to be ordered. The conditions are described below. Once it is determined another order-multiple is required, then the IRQ allotted to the chosen source is incremented, all further checking to see if another order multiple is required is bypassed and the store order round process is finished.

Effect of Order Pack and Normal Rounding Rules

When the total IRQ is less than one pack-size, Order Pack and Normal rounding rules are used to determine the order quantity. The decision of whether to order a pack-size is based upon the user-specified rounding method.

User Specified Rounding Method	Action Taken
Normal	Use Normal rounding rules.
Order Pack	Add pack-size if NI < SS.

If the instruction is Use Normal Rounding Rules, rounding is done according to the rules described on the sections below.

If the resulting action is Add pack-size if NI < SS, the IRQ is equal to one pack-size if the Net Inventory (NI) position is less than the Safety Stock (SS). No further rounding is needed. If the NI is not less than the SS, rounding is done according to the rules described on the sections below.

Effect of Freshness Flag on Rounding

The user-specified Freshness Flag is used to determine whether or not a store receipt alone must cover all customer orders over the review time.

Freshness Flag is True

Setting the Freshness Flag to True ensures that a store receipt is generated for at least the customer orders over the review time (CORT) regardless of the net inventory. In other words, the IRQ alone must be able to meet CORT. This ensures that the CORT is met by the freshest possible inventory.

Freshness Flag is False

If the freshness indicator is not on, the store's inventory is taken into account. If pack rounding the IRQ down does not cover the customer orders over the review time, the IRQ is pack rounded up.

Effects of Store Rounding Threshold on Rounding

The use of a user-maintained rounding threshold (RTH) indicates the percentage of an order multiple at or above which ordering an additional order-multiple should be considered. If the remaining order quantity as a proportion of an order multiple is greater than or equal to the rounding threshold, then an additional order multiple is required if any of the following conditions are met:

If product life is not relevant for the SKU.

If the total inventory at the store after rounding up does not exceed the calculated MSQ. Total inventory is calculated as the NI position plus the IRQ allotted to the source plus another Order Multiple.

If the total inventory at the store after rounding up (as calculated in the previous step) exceeds the calculated MSQ but the amount over MSQ as a proportion of an Order Multiple does not exceed the user-specified acceptable loss percentage (SAL). SAL represents the percentage of an order-multiple the user is prepared to lose in order to meet sales.

Pallet Rounding

The IRQ is now rounded to pallets. Pallet rounding only occurs if Cases per Pallet (CPP), which represents the Pallet Multiple associated with the Order Multiple, is greater than zero (0). Cases per Pallet are converted to Units per Pallet (UPP) by multiplying the CPP value by the Order Multiple.

Next the IRQ is calculated as a proportion a pallet. If the proportion a pallet is greater than or equal to a user-defined Percent of Pallet (POP) threshold, the IRQ is rounded up to the next whole pallet number if either one of two further conditions are met:

If product life is not relevant for the SKU.
If rounding the order up to the next whole pallet means that the subsequent inventory position in the store does not exceed the MSQ.

Warehouse Order Rounding

The following text lists the conditions under which another order-multiple is required:

If as a result of not ordering another order multiple, the inventory position at the warehouse would be less than the forecast demand over the review time, then another order multiple is ordered from the chosen source.
The user maintains a Safety Stock Threshold (SST) which represents the minimum percentage of the calculated safety stock (SS) that must be ordered when performing rounding. The calculation (1 - SST) therefore represents the maximum percentage of calculated safety stock that can be lost through rounding. Multiplying (1-SST) by the calculated safety stock gives the maximum number of units that may be lost through rounding. If the remaining Raw Ideal Receipt Quantity exceeds the maximum number of units that may be lost through rounding, then another order multiple must be ordered from the selected source.
The user maintains a rounding threshold (Rth) that represents the percentage of an order multiple on or above which rounding up to the next whole order multiple should take place. If the remaining Raw Ideal Receipt Quantity as a percentage of the order multiple is greater than or equal to the rounding threshold, then another order multiple must be ordered from the selected source.

Inventory Capping

When store space is at a premium, SKU Caps can be set to limit the IRQ.

Space (capping) constraints are applied to the Store IRQ for any SKU for which the Capping Flag is set to True.

The Store IRQ for a SKU on a given ATP day is calculated and capped by the Available Space on the ATP day.

Reconciliation

In the replenishment process, a receipt plan was generated for all stores to satisfy their future inventory needs. The replenishment receipt plan is unconstrained; it assumes the inventory at the source is unlimited. In reality, the inventory at the source is limited and may not meet the demand from the store. If the available inventory at a source is insufficient to meet the needs of all stores being supplied from that source, the shortfall reconciliation algorithm is run to distribute the limited inventory.

Reconciliation is source-centric, looking at the current inventory at the source, the aggregated demand from the stores and any other locations for that source, and expected receipts to project the available inventory position in the future. Using the available inventory reconciliation allocates shipments to stores and warehouses based on their priority and need. The raw need is used, and rounding and inventory capping decisions are reapplied to the quantities being allotted.

Reconciliation occurs during the Fixed Period (See Fixed Period definition) for all SKU packs, and in a limited set of circumstances post Fixed Period. Circumstances benefiting from post Fixed Period reconciliation are those that push excess stock, require prioritizing SKUs or packs, or inflict inventory constraints due to supply chain disruptions or end-of-life such as:

Stockless
Temporarily un-orderable
Event SKUs
Pack Change Event
Discontinuation Date
Stop Receiving Date

Two methods of reconciliation are available for a SKU. Day-on-Day Reconciliation allocates inventory based on orders that can be shipped that day. Future demand to be shipped at a later date is not considered. This method ensures that any available inventory shipped that day is in stores as early as possible.

Reconciliation over Time allocates inventory based on orders that can be shipped through the time of the next planned delivery to the source. The need over this time period (reconciliation period) is distributed across all stores. This method allows stores that would be shipped on a later date within the reconciliation period to receive a share of the available source inventory. In essence, the need for all stores for a SKU sharing the same source is respected over the entire period for the SKU/Source.

The reconciliation period for Day-on-Day Reconciliation is one day.

Reconciliation of Formal Packs versus Informal Packs

On a SKU by SKU basis the definition of a SKU's packs are either formal or informal. When a SKU is identified as having formal packs this means that the inventory of each pack is tracked through the supply chain.

In contrast informal packs identify a multiple of the SKU contained within the pack but inventory within the supply chain cannot be distinguished from one pack to another. As a result, inventory is only understood as a total number of units of the SKU rather than a number of packs.

The availability of pack-specific inventory plays a crucial role in reconciling the store orders. When pack inventory is not available, meaning the SKU has informal packs, the warehouse inventory is divvied out to stores using the store's ordering pack size to make rounding decisions. If all stores are uniformly ordering the same pack size then there is no discrepancy between the resulting plan and execution. Where the warehouse holds many pack sizes as a result of pack breaking, special purchases, promotional pack changes, produce packs, etc. there is greater opportunity for discrepancies. This is addressed with Pack-aware reconciliation which leverages Formal Packs to understand the makeup of warehouse inventory and the warehouse's pack breaking capability.

The warehouse only orders a single pack at any given time. As a result a few simple rules are applied to how the packs are shared:

Packs that are breakable are shared with all stores ordering a pack or any of its components (inners, eaches).
When a pack is not breakable, or an inner component of a breakable pack it is assumed that all stores are ordering the same pack therefore all stores have a shot at receiving the pack.
Packs are first set aside for Holdback, and Back Orders in the order of: pack to exhaust, then largest pack to smallest.

The same rules apply when a Pack Change Event has been defined, however all stores that order either the old pack or the new pack, and which are not excluded from the event have a shot at receiving a pack or its components.

When distributing packs that are breakable, or a component of a breakable pack the ordering pack size of the store on the delivery date is used to make rounding decisions about the allotted quantity. If the ordering pack differs from the pack being distributed because of a pack change then the appropriate old or new pack size is used in rounding decisions.

Total Demand and Available Inventory

The following sections describe Total Demand and Available Inventory.

Calculate Total Demand

Total demand at the source is calculated over the reconciliation period under consideration. Total demand at a source for a given reconciliation period is the aggregate of all receipts to the destinations from the source during the same period.

Total demand at the source is calculated as the aggregated demand from all receiving locations (stores and warehouses) from the current day through the last day of the reconciliation period.

Calculate Available Inventory at Source

Available Inventory is calculated over the reconciliation period under consideration. Available Inventory is an assessment of total availability at the source less quantity already committed to destinations that originate from that source.

When a source warehouse has External Demand the Prioritize External Demand Flag is used to determine when to allot available inventory to external demand.

When external demand is prioritized first the available inventory is first allotted to external demand before destinations.

When not prioritized, the available inventory is allotted to warehouse destinations and any remaining is given to external demand.

Allocation Process

Once it is determined that reconciliation is required, the process uses a priority matrix to determine each destinations allocation. Warehouses are fully reconciled before stores. While the priority matrices have similar elements warehouse destinations are prioritized ahead of stores. Further all warehouses have the same priority.

Each shipment was created to fulfill a destination's need. Need may have the following allocation boundary elements:

Customer Orders over Time (CORT)
Minimum Sales Stock (MSS)
Receipt Point (RP)
Receive Up to Level (RUTL)
Maximum Sellable Quantity (MSQ)

Store Reconciliation Matrix Process Sequence

Note:

MSQ is included in the Allocation Process because it is an allocation boundary. The MSQ is a boundary used for stockless items; therefore it are not included in the logic for the shortfall reconciliation.

In addition, each store is assigned a store priority. The store priorities include:

Super High
High
Normal

Store Priority	Allocation Boundary
Store Priority	CORT	MSS	RP	RUTL
Super High	1	4	5	6
High	2	7	9	11
Normal	3	8	10	12

The sequence process defines the order of priority of the combination of store priority and allocation boundary. For example, available inventory is first used to fulfill customer orders over review time (CORT) for super high stores (number 1 in the preceding example). If inventory remains, CORT for high stores is fulfilled. Next CORT for normal stores is processed, then minimum sales stock for super high stores, and so on.

The shortage priority matrix illustrates the priority order in which destinations served by a given source have to be allocated inventory in the event of a shortage.
Stores are not allocated inventory individually—they are allocated inventory as groups. Each store is assigned to a group based on its priority. For example, all Super High stores are grouped together for processing.
The user defined matrix that dictates boundary priority extends to warehouses. However, given that all warehouses have the same priority the allocation boundaries relevant to warehouses should be used in order of smallest to largest.

Determine Active Zone

The active zone represents the first allocation boundary that cannot be met for all destinations in a group using the sources available inventory.

This active zone process is applied to all shipments over the reconciliation period. The active zone is the cell of the priority matrix to which available inventory cannot completely fulfill. For example, if available inventory was sufficient to fill up to the minimum sales stock (MSS) for all high priority stores (zone 7) across the reconcile period, but not the MSS for normal priority stores (zone 8) across the reconciliation period, zone 8 would represent the active zone. All inventory requirements of zone 1 to zone 7 would be fulfilled with available inventory. Residual inventory would be watershed across the stores in zone 8 across the reconciliation period.

Use Watershed Algorithm for Allocation

While the watershed algorithm is complex, the basic premise is to bring all destinations in the priority group up to a target ratio. The Target Stock Ratio is calculated by dividing the source's total available by its total need.

One way to think of this procedure is through an analogy of water filling a container. The fill line of the container represents 100% of need for all the locations, and the quantity of water we have represents the stock available. By pouring all of our water into the container the water level naturally reaches our Target Stock Ratio. If we imagine the bottom of the container to be partitioned to represent our different destinations, then the water above each partition represents each location's stock ratio as well. The water naturally levels itself to maintain a consistent stock ratio across all the destinations.

Figure 11-1 Watershed Example

In this example, our target ratio is 64%. The purple bars on the chart represent the amount of inventory available at each store. The blue bars represent available inventory from the source allocated to the stores. In this case, only 6 of the stores are allocated additional inventory.

Substitution

Substitution is the act of sending one SKU in place of another when the demand cannot be met. This is done with the expectation that in the event of a stock out at least some shoppers will select an alternative substituted SKU instead of going without.

A percentage of sales, known as the Demand Transfer Percentage, that is believed to transfer to the other SKUs in the event of a stock out are loaded into AIP for use in the substitution process.

Substitution is performed after all other attempts to meet demand have been exhausted. Substitutions of this nature are not planned into the future. They are only performed in the last batch run of a SKU prior to release of the store order.

This process will attempt substitutions at stores on days where a shortage is identified and the source of the shorted sku/store ships the inventory today. A source may ship inventory today because of the natural lead time of the shortage day, or because the source is stockless today and must meet all demand in the reconciliation period today. For an ATP day to be considered for substitution, all of the following conditions should be met:

Inventory is less than the Minimum Sales Stock (MSS) boundary. The days meeting this condition have already had their need-for-substitute flag set to True by the reconciliation process.
The ATP day ships today or the ADD of the ATP day ships today.
The SKU/store releases in the current release wave (where no wave is specifically defined, release wave is Overnight) or is not Re-planned prior to release intra-day.

The process steps over each SKU/store/ATP day that has stock outs (they need substitution) and transfers the day specific unsatisfied demand to the substitute SKUs using the appropriate Demand Transference percentages. Customer Orders and Back Orders are never transferred to another SKU. Once all SKU/store/ATP days are completed the transferred demand is summed for the review time(s) of each substitute SKU/store. The review time demand transference is applied to certain boundaries to increase the targeted inventory levels.

After all transferred demand is calculated the source of the SKUs that have received transferred demand are identified. For those source/SKUs the watershed algorithm is applied to determine if the inventory in (or expected in) the store meets the increased inventory boundaries and share the source available inventory equitably where there is now additional store need. After any needed inventory is transferred from the warehouse to the store the final substitute quantity that could actually be met by the available inventory is recorded so that the store projected inventory is reflective of the demand increase expected.