From SAP Advanced Planning and Optimization (APO) safety stock planning, through SMART Ops, Enterprise Inventory and Service EIS on to SAP Integrated Business Planning (IBP) for inventory (IBP-I), safety stock planning has evolved from isolated single stage to multi-stage inventory optimization. In practice this means moving from planning safety stock at an individual location to optimizing safety stock for an entire demand stream. The value is now in network optimization as opposed to planning at the individual warehouse.
SAP IBP-I is an integrated multi-stage inventory optimization solution that enables seamless Sales Inventory and Operations Planning (SIOP) and provides network consensus safety stocks for supply planning. This sounds wonderful in theory, however, we must be aware of the implementation hurdles to success.
1. Supply chain synergy is a must
IBP-I is born out of SAP EIS, its predecessor being SMART Ops acquired by SAP in 2008. The principle is to generate an optimal target inventory position (TIP) per stocking point, driven by a target service level (TSL) for a customer, or group of customers at the lowest cost possible across a supply chain. The TIP includes your safety stock and cycle stock to make up your core on-hand stock and pipeline stock. It is calculated using a mixed integer programming optimization (MIPO) stochastic algorithm.
MIPO supports multi-stage inventory calculations for complex multi-strand supply chains. A supply chain view is taken considering the demand and supply relationship of an edge within a path where an edge defines any link between 2 nodes of the supply chain and a path defines the route from supply point to the customer group. During MIPO demand is propagated through the supply chain paths. MIPO determines the optimal internal service levels based on a number of parameters including inventory holding costs, lead time and lot sizes as well as uncertainty data and then applies this to the nodes in a path to calculate the TIP at each stocking point in the supply chain.
There is no halfway house, you must have a holistic view of the supply chain. A stocking point in the network cannot decide to adopt its own policy. It must adhere to the validated output of the run to ensure there is inventory harmony across the multiple stages otherwise you risk carrying too much or too little stock. For inventory optimization to be effective the output must be adopted by all nodes in the supply chain path.
"For Inventory optimization to be effective the output must be adopted by all nodes in the supply chain path"
2. Understand your scope of integration
It is important to understand the theory, but also to focus on what is actually needed in practice to realize the benefits of inventory optimization. Do not underestimate the integration effort. A substantial benefit of SAP's IBP solution is the ability to build a Sales Inventory and Operations Plan (SIOP). In theory you are able to take the output from your multi stage inventory optimized run, and apply this to both your Sales and Operations plan and your Supply plan. Much like time series key figures in APO, data is not transferable between planning areas within IBP. Our team have used HCI to extract data out of the inventory planning area into the SOP planning area. In future versions where a single planning area should be a reality, we expect the integration to be even more straightforward.
If you are an APO user and you want to transfer the resulting recommended safety stock for use in your midterm SNP plan, the integration challenge is akin to integrating Smart Ops or EIS solutions to APO or ECC. With EIS, SAP developed a material master add-on that enables key fields required to drive EIS in ECC. A series of custom business application programming interfaces (BAPI's) are then provided to support the extract of this data to EIS.
As a business, you must decide on the type of integration required and the benefit expected in the short term vs the longer term view. Having been deeply involved from design through build for an EIS, APO and ECC automated integration project at a FTSE 100 client, the integration challenge is substantial but not insurmountable. Process Orchestration (PO) was central to the integration design. In the case of IBP, HANA Cloud Integration (HCI) is central to mapping the right data to be extracted from external systems and transferred to IBP. The solution is robust however does require a good deal of configuration, ABAP program generation, mapping and depending on the type of data, uses APO’s planning area data sources.
"Do not underestimate the integration effort"
3. Quantify the benefits of time varying safety stock
A common practice I have frequently seen is the tendency to apply a static rule in inventory parameter settings. The assumption is because a location exists a certain distance from its source of supply, you will, irrelevant of demand, replenish a full truck load or pallet load of the product when replenishment is triggered. Alternatively, you assume you will always carry 30 days of forward inventory cover irrelevant of a product being a runner, repeater or stranger. Agreed there is an element of generalization however the point is that one size does not fit all. In almost all cases this leads to higher than necessary inventory costs or a tendency for shortages, stock outs and lost sales. Within APO, safety stock planning has rarely been as meticulous as it should and this could be as a result of the simplistic planning approaches used.
If you are an existing APO user, it is likely you are generating safety stocks in one of a number of the following ways.
Using SNP safety stock planning to generate a single statistically determined safety value that is written into the product master and used by your planning engine.
- Using a day’s coverage method, such as a safety days’ supply or time as your forward cover of forecast demand
- Or any combination of the above where a static safety stock is used as a failsafe if lower than your forward cover.
Ideally in the coverage method you will write the results into the planning book so that planner adjustments or exceptions above max or below min can be corrected. The functionality in itself is simple and straightforward however it is isolated by location and product and standardized by coverage duration and batch size.
The focus on IBP for inventory’s optimization algorithm is on ensuring that the internal service level is met at each location driven by your forecast within the constraints of cost, replenishment lead time, lot size and the uncertainty of supply and demand. The aim is to have the right balance of product to achieve the service level at that location and in so doing safety stock must be time varying due to the other hard constraints imposed. With fluctuating demand and fixed lead times and batch sizes safety stock acts like a car’s suspension. During times of low demand safety stock is reduced, and during higher demand scenarios safety stock is buffered against uncertainty. This in itself lays the ground work for a strong business case when compared with current practice.
"Within SAP APO, safety stock planning has rarely been as meticulous as it should"
4. Identify the opportunities to improve planner productivity
Planner buy in is key to ensuring success. However, it should not just be your supply planning team, a business user team that is able to understand wider business supply targets and cost reduction goals within the context of how it relates to inventory is fundamental to ensure real business benefit can be delivered. The team must understand the process drivers through IBP for inventory that will drive value. Apart from the foundational of optimised time varying safety stocks, there are lots of highlights which come as standard with IBP that when used in the context of inventory will improve planner productivity. For example, the reduction in short term firefighting paves the way for value adding activities such as; what if analysis for stock builds, production shut downs, impacts of seasonality and promotions. Planners can also take a monitoring role of their respective supply chains to ensure objectives are being met and to hold each other accountable against service level expectations in the supply chain.
The improvements over APO and EIS in network visualisation and front end analytics are enough to entice any planner to the tool. The familiarity of excel replaces the rigidity of APO interactive planning books. The picture below of the IBP dashboard shows selected charts that a planner is able to interact with as a home page.
Through the Excel front end, you are able to simulate or execute inventory optimisation for a given supply chain network and review the results graphically.
Excel views can be personalised and shared. Having worked with numerous planners over the years, I believe this is a welcome development.
"The team must understand the process drivers through IBP for inventory that will drive value"
5. Be clear on available functionality
IBP for inventory is a complete rewrite of EIS, with improvements made to leverage the phenomenal performance benefits of HANA. That said, Inventory Optimization is growing in use but is not as widespread as one would expect. Implementation expertise is hard to find in comparison to SCM APO. IBP for inventory is delivered with the ability to run single-stage algorithm calculations, as well as multi-stage using MIPO, with the added functionality to check expected demand loss of a plan. This allows an organisation to improve on its current solution without going to a multi-stage optimization solution. Potentially what it also enables is a solution where by a fractal supply chain can adopt differing solutions depending on complexity and cost saving.
Service Level Optimization (SLO) that is available in EIS but not in IBP for inventory is planned for a future release. There is also some rationalization from EIS into IBP for inventory, for example periods since last review (PSLR), and periods between shipments (PBS) are dropped and only Periods Between Review (PBR) remain. The underlying configuration of planning areas and attributes with planning levels and master data types is developed to improve performance with HANA. So far we have not seen Demand or Supply Intelligence modules which in EIS are used to determine your coefficient of variance for uncertainty modelling. Without this the lead time definitions are business input based on existing practices rather than determined by the system. We will address the pros and cons of intelligence modules at a later time.
From a business perspective it is important to manage expectations in line with available functionality to determine what can be done and what is practical within different regions and different product channels and categories. This needs careful planning from the outset. With each function and solution comes a set of possible benefits. It is important therefore to understand your current performance and to create a basis for a comparison to be able to measure the impact of the solution.
There are both technical and business challenges to be considered prior to embarking on your IBP for inventory journey. A strong business case is one that quantifies the impact of optimised time varying safety stocks throughout the supply chain and improvements in planner productivity on customer service levels, inventory as a percentage of working capital and effective asset utilisation.
Click here for a full list of Olivehorse IBP blogs.