Keeping Ventana Models Connected

Client expectations about how changes in one factor will affect other factors are documented using Vensim® Reality Check®. Some Reality Check examples:>

• "Customers will only share information if they believe their privacy is secure."
• "When service quality drops, our most profitable customers leave faster than our less profitable customers."
• "Past a certain point, overstaffing will decrease workforce morale and service quality."

Vensim's patented Reality Check technology records these statements of expected behavior in model libraries, in a format that allows automatic testing of the model against each statement. Users can see visual output confirming model conformance to expectations. If the model fails a Reality Check, it is an immediate warning that there are unexpected effects in the model relationships, indicating that the model, the expectations, or both must be revised.

Reality Check effectively connects the model to organizational knowledge in two ways. First, it provides an easy way for people to contribute their expertise without having to know anything about the technical details of the model. This knowledge is the basis for drafting initial model relationships, and is then used for ongoing checks of model behavior. Second, showing that the model behavior conforms to Reality Check expectations is a powerful demonstration that the model's internal dynamics work in accordance with the client's understanding of their business. This demonstration is a valuable aid in establishing model credibility. The set of Reality Check statements can continue to grow with model use, powering ever-increasing model reliability, providing an enduring archive of organizational knowledge, and facilitating ongoing testing of the archive against new data and hypotheses.

Models use historical data in two ways. Some data are used as direct input, for example, advertising spending data. Other data are compared to model output--for example, product sales data--both to help calibrate the model, and also as one of many tests of model reliability. Typically, Ventana models draw on data from multiple sources in multiple formats.

Ventana offers a wide range of data management options to collect incoming data, organize, store, and make them available to the model, and to archive output to benchmark model performance. Small amounts of data that do not require frequent updates are often best collected by hand in spreadsheets, which are then read by the model. Larger amounts of data can be automatically drawn from one or more existing databases, or compiled in a dedicated database. Ventana customizes data management procedures to suit client needs, balancing initial development cost against the ongoing cost of maintaining and using data to find the most productive option.

When required, Ventana models can handle massive amounts of information. Connecting data to Ventana models makes it feasible and efficient to inspect, preserve, and use large data sets productively when traditional database applications cannot. For an example, please read about Ventana's spares logistics application. >
>

A user interface for the model and a tailored training program are designed together for a common purpose: producing a system for users that

• highlights and clarifies most cared about results,
• makes common tasks comfortable and simple, and
• provides flexibility and analytical capability appropriate to the business need and users' backgrounds.

Ventana provides a number of options for model interfaces and training.

User interface options:>

• Users can operate models directly in the Vensim modeling environment
• A Ventana custom graphical user interface eases common tasks, lowers training requirements, and emphasizes most important results. Graphical user interfaces (GUIs) and web capability can be created from Vensim-based tools, including Venapp and Sable interface development software.
• Vensim simulations can be controlled from other applications, such as Microsoft Excel™
• For specialized needs, Ventana can create custom, freestanding applications.

Training options:

• Ventana regularly offers basic and advanced courses in using the Vensim modeling environment. Please see the Vensim web site for more information.
• Ventana can provide custom, on-site training focused on the client's model or on general use of Vensim.
• Ventana can also provide a "train the trainer" course, to help initiate in-house training programs taught by client personnel.

Ventana helps each client choose the most cost-effective and productive combination of interface and training.

Getting full advantage from Ventana models is a matter of supporting and coordinating people, and how they spend their time. Since the client knows best how to deploy its people in the context of their other responsibilities, Ventana helps the client to design their own business process for the model. The business process: >

• Schedules interactions between decision makers and the model, either directly or through designated model users.
• Provides time and tools for people to interpret and communicate model results.
• Facilitates upkeep of data sets and minor model maintenance if required.

The amount of time required for this depends on the degree of automation built into the interface and data management system, and on the nature of the decisions being supported. Ventana helps clients to determine their specific requirements. Clients have developed a wide variety of means to support smarter decisions with models. Some common arrangements include:

• Executive strategy application: Executives examine their options using a decision support tool on their own personal computers, using an intuitive interface requiring little training time, and automated data management supported by an analyst. (Example: The CEO Strategy Flightdeck for financial institutions)
• Model expert: Decision makers communicate regularly with an expert model user who generates, interprets, and explains model results.
• Outsourcing: A consultant uses model and reports results and implications to the executive team.
• Embedded algorithm: A Ventana model is used to develop and test decision rules for automated decisions, for example, logistics rules for re-ordering, stocking levels, part transfers, and the like. These embedded algorithms may be updated using a strategic model, for example, a model of the overall supply chain system. >

Drawing on past experience, Ventana helps clients define the business process best suited to their own needs and culture. > Ventana helps clients make coordinated choices for knowledge management, data management, interface, training, and business process, to put in place solid tools and procedures for sustained performance improvement.