Super Models: The Best of BPM Analytics
When data warehousing and business intelligence technologies emerged years ago, business modeling was a key component of business intelligence software. Since then, however, the vital process of business modeling has been largely overshadowed as an area of focus, for both vendors and corporate implementers of business intelligence, by methods of reporting, dashboards, and other forms of data presentation.
Article Tools
Resource Center
Access white papers, product demos, and presentations from companies whose reputations have been built on helping BPM practitioners get the most from initiatives.
- BPM 101: Selecting a Business Performance Management Vendor" -- new white paper from BPM Partners
- "The Finance Challenge of Aligning the Business With Strategic Goals," a podcast featuring Palladium Group's Phillip Peck
- Ventana Research white paper "Decision-Making and Performance: Improving Essential Business Analytics and Technologies"
- “XBRL at a Glance,” white paper from XBRL US
advertisement
Meanwhile, the business community has continued to use spreadsheets for modeling because tools dedicated to this critical practice have been unable to match spreadsheets' ease of use or flexibility. Specifically, the spreadsheet has been almost irreplaceable for modeling because of its expressiveness -- in other words, because a model can be rendered in a spreadsheet almost as easily as it can be conceived by business and finance managers. Of course, spreadsheets have the same drawbacks when they serve as a modeling platform that they have when used in budgeting and other areas of business performance management (BPM): They are hard to maintain and share, and documenting changes is nearly impossible.
Creators of business models have never had an ideal solution to turn to, but that's beginning to change. Some BPM systems are beginning to combine the flexibility of spreadsheets with a centralized data repository.
Why Spreadsheets Aren't Sufficient
All companies use some form of modeling. Business models help managers forecast the effects on the organization of possible changes in pricing, compensation, allocations, and yield management, among other factors. But most of them are tacit and implied; they don't follow a formalized methodology that is shared among business managers throughout the company.
When a model is hidden, or at least is not publicly recorded, it is not available for interactive discussion among various people in the company. The organization may lose knowledge of the model if its owner changes jobs. In addition, an unshared model results in a variety of processes which are not standardized across the organization. A business model in one division of the company may be vastly different from those in other divisions, so its results may differ substantially from those of the other groups' models. Although tacit business models work in a limited way to help individual department managers prepare for the short term, they are not efficient and do not make long-term business sense.
A more strategic approach incorporates straightforward modeling functionality, including the expressiveness most commonly associated with spreadsheets, into a comprehensive BPM software system. In general, the benefits of incorporating a business model into a BPM framework result from the software's architecture, which separates physical data models (i.e., the technical aspects of how the data is collected and how and where it is stored) from semantic models (those in which the logic and relationships among the data points are expressed). BPM software is capable of providing the tools businesspeople need to manipulate data visually and in terms they understand so that they can concern themselves only with the meaning of the data -- not its structure, location, or format. This shift toward technology that expresses the rules of business, not just the metrics, will elevate the role of modeling and make it a top-of-mind topic for finance, business, and IT executives.
What a Good Model Offers
Business models vary from the simple (those from which templates are often generated) to the highly complicated. A basic set of data may be the key ingredient to a simple model; data input and arithmetic may be all the functionality that the model needs. Conversely, complex business models may use business rules, convoluted formulas, and sophisticated calculations to handle trending, forecasting, analysis of transaction patterns, projections, determinations of risk, and more.
Regardless of its complexity, a model is ultimately going to be used by line-of-business managers and financial analysts, so these people should be able to change it. When logic is added to a model, it becomes procedural -- that is to say, the model is then programmed to follow a specific set of instructions. Mixing calculations and logic can yield a very potent tool. Unfortunately, procedural models are difficult to develop with most applications available today, and they are even more difficult to maintain and modify. Working with procedural models in most business intelligence packages requires interaction with the system at a coding or scripting level, something outside the skill set of the average business-level user. Businesspeople -- even the "power users" -- should not be expected to know software engineering techniques. Asking them to do so is neither realistic nor economically efficient.
