Other Requirements for Fully Functional Models

A modeling tool which includes all of these basic traits will provide functionality that helps businesses operate better. Still, a few additional attributes can make business models even more valuable to companies by making them as easy as possible for businesspeople to use.

One of these features is the look and feel of a spreadsheet. Spreadsheets are widely used and easily deployed. Plus, they deliver a number of benefits specific to modeling. They can be used to build models anytime the need arises, then to easily integrate those models with multidimensional OLAP tools. This combination enables the user to create a model in a spreadsheet, then replicate it across data dimensions. Macros can be built without writing code, so the spreadsheet can mimic what an analyst might do manually. Spreadsheets are expressive and enable people with limited training to develop a wide range of business models.

For spreadsheets to facilitate truly effective modeling, they must work in a collaborative environment. Thus, in their interactions with the underlying database, they must have both publishing (also known as "read") and revision (known as "write") functionality. The read/write approach allows for dynamic adaptations and ensures that the models are updated as appropriate. A read-only approach to a spreadsheet-based model limits both its usefulness as a modeling tool and its ability to be modified by businesspeople.

Another attribute that takes a modeling tool above and beyond the mainstream is simplicity. Once business models are in use in multiple divisions of an organization, propagating companywide changes to the models can be a challenge. Additionally, implementing changes introduces the problem of version control. To maintain simplicity, and to live up to the promise of common business models, organizations should adopt a uniform rules statement that drives the values (such as planning values) of its entire multidimensional modeling environment. With a single rules engine, if a change is made in any business rule or planning assumption, the values are automatically recalculated and can be viewed instantaneously.

Along the same lines, modeling software needs to be comprehensive and intelligible. As IT departments learned in the mid to late 1990s, documentation does not always keep up with the application or application coding. The less automated the process of modeling is, the greater the chance that changes will be made to models without clear documentation of the reasons for the changes or the assumptions underlying them. While this doesn't necessarily create the version-control nightmares that one-off models do, it does mean that statements, values, and assumptions could inadvertently be "hidden" in a model. Increasing the level of automation in a modeling tool -- through self-documentation and audit logs -- is valuable because then the tool can provide documentation, clarity, and traceability of changes to individual models.

Security in business-modeling software should be no different from that of an enterprise file server: File access should be granted based on a person's role, responsibility, and job function. This type of security measure rarely exists in the realm of financial data and spreadsheets; sometimes finance must choose between everything or nothing being locked down. But a modeling tool that incorporates the idea of role-based security -- and for spreadsheets, cell-level security -- allows organizations to more fully exploit the benefits of modeling.

Finally, a super modeling tool must score well in the areas of transparency and accuracy. These are two of the biggest buzzwords in business today. Just as in accounting, they hold relevance for modeling. With modeling, once the metadata is established, the system automatically understands, say, the difference between assets and expenses; provides accurate computation; and enables users to transparently understand the logic in the model.