SAP APO: Will it Fill the Gap?

Product Background

Announced over two years ago, SAP's Advanced Planner and Optimizer (APO) takes aim at the burgeoning supply chain management market, where it will compete head-to-head with offerings from established best-of-breed vendors like i2 and Manugistics. In the mid-nineties, responding to pressure from a number of Fortune 50 companies, SAP devised the Logistics Optimization Interface (LOI), a project that enabled supply chain vendors like Chesapeake Decision Sciences, Inc., i2 Technologies, Inc., and Numetrix Limited to integrate their applications to R/3. While other enterprise application providers like PeopleSoft and JDEdwards have acquired many of these smaller companies (see Table 1), SAP used the collaboration as a prelude to an internally developed offering. Reluctant to acknowledge any input from former partnerships, SAP credits the breadth, depth, and relevance of functionality within APO to their association with the Supply Chain Customer Advisory Council, a group of companies who co-develop requirements and provide ongoing feedback to SAP's development team. Now, with a massive 190 pilot installations underway in North America, Europe, and Asia, SAP is pushing furiously to complete development on the core functionality of APO.

Table 1 Markets Converge: SCM Acquisitions by Some Leading ERP Vendors

Product Strategy and Trajectory

SAP is targeting APO primarily at the electronics and high tech (EHT), consumer products, metal and paper, automotive, pharmaceuticals and chemicals industries. APO promises broad functionality, real-time access to transaction data and seamless integration with SAP R/3. Although the first companies to acquire the new package will likely be current R/3 users, SAP plans to offer APO as a stand-alone supply chain management (SCM) application that can be integrated to legacy and third party ERP systems. Table 2 lists some of the major shortcomings of R/3 and explains how APO is positioned to solve them and bring additional benefits.

Table 2 SAP APO: Fixing a Hole

Product Strengths

• Intuitive GUI with menu-driven modeling capability: APO enables detailed modeling capability and configuration directly through its user interface. This provides an advantage over packages like i2's Rhythm, which requires that users construct models and set key parameters in separate files, then load them into the application.

• User flexibility: Modules such as SNP/D offer a variety of advanced algorithms and can perform cost-based optimization, an option that is often lacking in competitive applications. Global ATP (Available-To-Promise) provides a highly configurable rule-based language for defining availability by product, customer, etc. that allows users to model a large variety of ATP schemes.

• Good integration between modules: Modeling constructs and rule application are consistent from module to module, enabling a more seamless type of integration than competitive supply chain management suites, most of which rely on acquired solutions with disparate technologies.

Product Challenges

• Data integration: Although much of the static data needed by APO is readily available in R/3 (e.g., bill of materials, product data), a significant amount of the data must be maintained separately from the ERP system. This presents an added maintenance burden that compromises the benefts of "seamless integration".

• Web-enablement: Thus far, the web-enabling features of some of the modules have not met expectations. For example, in the DP module, there is no capability to share forecast information via e-mail or generate reports on the web.

• Dynamic memory size limitation: LiveCache, APO's dynamic, memory-resident database, allows for a maximum size of 3 GB. For large implementations in high-volume industries, such as PC or semiconductor, this can pose a severe restriction on scope and/or present significant challenges in data segmentation.

• Some modules are short on graphics that would facilitate the evaluation of planning results. For example, PP/DS relies on Gantt charts with task lists to illustrate factory-level scheduling results. Though it may accurately reflect the plan, this technique cannot provide an aggregated view of production requirements.

Vendor Predictions

• Though slow to emerge from its incubation phase, APO has all the makings of a market-leading product. Backed by a well-funded development machine and strong direct sales force with a large customer base, APO should begin generating competition for the best-of-breed SCM vendors by mid-year 2000. (70% probability)

• In keeping with its goal to penetrate the mid-market, SAP will likely offer a stripped down version of APO within the next 12 months that will include a smaller version of the Business Information Warehouse (BW) and liveCache. (60% probability)

• Although SAP maintains that APO will be sold as a stand-alone SCM application, such strategy contradicts their traditional "big-bang" approach to marketing. It is doubtful that license revenues from individual APO sales will exceed 10% of total corporate revenue over the next 24-36 months. (70% probability)

Vendor Recommendations

• Utilize pilot implementations to draft industry-specific templates for APO in targeted markets, such as EHT and consumer products. Templates can greatly reduce the time and cost of implementions by encapsulating relevant business processes within a framework tailored specifically to an industry's general requirements.

• Continue development on utilities for mass data load and maintenance. Although SAP expects most data to originate in R/3 and legacy systems, mass data entry can greatly aid implementation teams in constructing and testing pilot models.

• Push to complete development of the core modules in APO version 2.0 (due in October 1999) and allocate application development resources to legacy and third party software integration issues.

User Recommendations

• Companies that have R/3 and who are now planning a foray into supply chain management may want to start with APO Demand Planning, as it is the most mature of the modules in functionality and integration to R/3.

• Until SAP can improve the efficiency of its optimization routine, APO users should judiciously confine the scope of their implementations to only the most critical items, business units and production facilities.

• Due to some overlap in functionality, R/3 users need to decide carefully how to best utilize features of R/3 and APO to avoid conflicts. For example, R/3's ATP functionality may be sufficient for companies who have never used ATP features before, but would benefit from APO's demand planning capabilities.

• Companies who need a stand-alone SCM system would do well to bypass APO in favor of established suites, such as those of i2 or Manugistics. An essential milestone for any new product, at least five live installations should exist prior to adding APO v2.0 to a short list.

• Companies that are more daring should seek large price breaks on new modules due to unproven performance. Currently, for version 1.1, the main modules in order of maturity are DP, SNP/D, ATP, and PP/DS.