The Promise of EPM
By John Hagerty and Dana Stiffler -- Supply Chain Management Review, 7/1/2003
Enterprise performance management (EPM) sounds simple: (1) define the metric, (2) set the goal, (3) monitor the process and measure performance, (4) send out an alert when a metric is out of whack, (5) take action to adjust process, and (6) repeat. EPM today is top of mind among business executives, and leading companies are beginning to achieve significant, quantifiable benefits from it. But getting even one line of business to the point where they actually see those benefits is a major undertaking. It requires large reserves of cash, discipline, and patience.
EPM is a robust business process that builds on the foundation of technologies and applications a company has already implemented. It typically incorporates business intelligence products, reporting tools, planning and budgeting applications, analytic applications, incentive management systems, portals, and scorecards along with data warehouse technology, data models, and integration software.
Users who've implemented EPM projects tell us that subpar data quality and architecture are the main obstacles that need to be overcome. More often than not, projects require a data warehouse at the front end. As a result, EPM initiatives are more complicated and expensive than expected.
Despite such obstacles, the most advanced companies are enjoying the initial fruits of performance management success, and they plan to keep investing in this area. AMR Research recently spoke with two dozen North American and European companies to dig into their EPM stories, identify the tools used and pitfalls encountered, and form a basis for others that are just starting down the performance management path.
Effort Often Exceeds Expectations
According to these companies, EPM initiatives, at the outset, often run up against a brick wall. The disparate systems and data that exist at a company usually mean that when an EPM tool is implemented, everything that comes out on the other side is rubbish. The GIGO rule—garbage in, garbage out—definitely applies. Addressing the problem is not a simple matter, as evidenced by the following examples:
-
Company A wanted to install supply chain performance management software but discovered that it had 17 disparate pools of supply chain data. A two-year data warehouse project ensued.
-
Planning to commence an ambitious performance management project, Company B used a single enterprise vendor but had several highly customized versions of its applications. Once again, a lengthy data warehouse project ensued.
None of the initiatives we looked at would make it past today's tough-minded CIOs, who demand small teams, fixed schedules, and a hard return on investment (ROI) up front. (And the fact that these projects take longer and cost more than anticipated indicates that business leaders—to be more precise, C-level business leaders—are driving the bus.)
The minimum initial project cost we encountered was $500,000, including hardware, software, and services. (Exhibit 1 shows the spectrum of EPM project costs for selected users.) Notably, projects at the lower end of the cost scale already had addressed their data concerns or were confining their initiative to a single line of business. Interestingly, users that selected their enterprise resource planning (ERP) vendor for performance management did not experience noticeably lower costs than users choosing the best-of-breed route.
For heterogeneous enterprise environments, customers reported that a best-of-breed vendor is the sensible choice. The very concept of enterprise software suites calls into question (at the present time, anyway) those vendors' commitment to unbiased interfaces with multiple applications and data sources.
For those with common ERP architectures, the enterprise vendors' offerings merit serious consideration. If the selection process is heavily influenced by IT professionals, the enterprise vendors' performance management modules are more likely to get the nod—even if they are not the unanimous choice from a business point of view. For instance, in companies where SAP dominates, that vendor's Strategic Enterprise Management suite requires only an incremental increase in IT resources compared to introducing a new vendor and technology architecture into the mix.
Do EPM and ROI Mix?
Pre-project assessments can highlight the softer benefits of EPM, but harder returns are tough to forecast. This is not a straightforward "fire 20 people" or outsourcing decision that instantly affects the balance sheet or P&L. It's more about shining a light in corners to see what's there and, more importantly, what you then choose to do about it. Performance management is a multistep process, not a single project: Users characterize it as implementation followed by "endless tinkering." The benefits generally go unrealized until a year or more into a project—and they're often soft at that. For example, while the organization may get visibility into performance, it's not able to act on that capability.
Despite these challenges, EPM initiatives are a priority for many business executives, and for good reason—they're working. In marked contrast to the grumbling we often hear about ERP implementations, not a single executive we spoke with regretted embarking on these projects. While returns may not be instantaneous, they are undeniable, as summarized in Exhibit 2.
Executive or board-level mandates to implement EPM projects take care of many ROI questions. One of the most ambitious (and expensive) projects we examined was justified by the decree of a single individual: "I want this thing, and I don't care what anyone says." He was the chairman and CEO of a Fortune 500 company.
Making it Happen
Business drives EPM, but IT needs to ride shotgun. The importance of the upfront data piece means that IT really does need to stand shoulder-to-shoulder with business on EPM implementations. Because these are long-term projects that require intimate knowledge of a company's strategy and operations, the likelihood of a third party being called in to engineer everything is slim. For the most part, companies are doing the implementation on their own, with some help from the vendor's professional services groups. Also, remember that depending on which strategy/tool is employed, performance management requires ongoing internal resources on both the business and IT sides.
EPM tools differ in their initial costs and implementation time. They also differ in terms of what it takes to support them. Prebuilt applications are quicker to implement and deliver a payback, but they are also more generic and less flexible in their analytics. For example, users focusing on a preconfigured tool for one piece of their supply chain—say, inventory—experienced quicker returns (although nothing resembling instant gratification). At the other end of the scale, certain analytics packages rely heavily on IT and specialist staff for their care and feeding.
With improved visibility into business and operations, users are able to identify problems—and, with luck, root causes. However, being able to react quickly and consistently remains a challenge. The policies and procedures that support this ability must be embedded in your EPM initiative; in other words, change management becomes a major requirement. For example, what if you need to reduce a certain product line by 50 percent ASAP, but you have long-term contracts in place with your suppliers? This is akin to being informed that a hurricane is headed for your house. Even though you know something about hurricanes—what they're made of, how they're formed, the damage they can do—this knowledge in and of itself isn't particularly helpful. It won't save your house. To effectively protect your home—or enterprise—you need to do the following:
-
Be ready. Use analytics as a catalyst for business process change within the company. Be ready to execute once you identify issues.
-
Start small. If data from a line of business is in relatively decent shape and key performance indicators can be identified, start there.
-
Synchronize implementation. If data warehousing is already underway in your company, implement performance management in tandem with that initiative.
-
Don't shortchange yourself. Particularly on the initial EPM project, use your very best people both on the business and IT sides. Do not rely on what are considered "surplus" resources for these initiatives.
-
Spend smart. Companies using global IT services firms are spending at the top end of the EPM budget range. To reduce project costs, use these services firms sparingly for tactical technology and/or domain expertise.
-
Proceed with caution. There is the potential for overconfidence among users that have already implemented a data warehouse or ERP system. While these experiences might give you a leg up on the EPM implementation, the related process challenges are still tricky.
View All On-Demand Webcasts
