How to Develop a Predictive Analysis Model for Specific Decisions

By Anasse Bari, Mohamed Chaouchi, Tommy Jung

A good model needs a prototype as proof of concept. To build a prototype for your predictive analytics model, start by defining a potential use case (a scenario drawn from the typical operations of your business) that illustrates the need for predictive analytics.

How to fit the predictive analysis model to the type of decision

Business decisions take diverse forms. As you undertake to build a predictive analytics model, you’re concerned with two main types of decision:

  • Strategic decisions are often not well defined, and focus only on the big picture. Strategic decisions have an impact on the long-term performance of your company — and correlate directly with your company’s mission and objectives. Senior managers are usually the ones who make strategic decisions.

  • Operational decisions focus on specific answers to existing problems and define specific future actions. Operational decisions usually don’t require a manager’s approval, although they generally have a standard set of guidelines for the decision-makers to reference.

The two classes of decisions require different predictive analytics models. For instance, the chief financial officer of a bank might use predictive analytics to gauge broad trends in the banking industry that require a company-wide response (strategic). A clerk in the same bank might use a predictive analytic model to determine credit-worthiness of a particular customer who is requesting a loan (operational).

With these two major types of decisions in mind, you can identify colleagues at your company who make either operational or strategic decisions. Then you can determine which type of decision is most in need of predictive analytics — and design an appropriate prototype for your model.

How to define the problem for the predictive model to address

The basic objective of a predictive analytics model is make your business do better. You may find that a prototype of your model will work best if you apply it to operational examples and then link those results to solving broad, high-level business issues. Here is a list of general steps to follow next:

  1. Focus on the operational decisions at your company that can have a major impact on a business process.

  2. Select those processes that have a direct effect on overall profitability and efficiency.

  3. Conduct one-to-one interviews with the decision-makers whose support you want to cultivate for your project. Ask about

    • The process they go through to make decisions.

    • The data they use to make decisions.

    • How predictive analytics would help them make the right decisions.

  4. Analyze the stories you gathered from the interviews, looking for insights that clearly define the problem you’re trying to solve.

  5. Pick one story that defines a problem of a small enough scope that your prototype model should be able to address it.

    For instance, suppose you interviewed a marketing specialist who is struggling to decide which customers to send a specific product ad. He or she has a limited budget for the ad campaign, and needs to have high confidence in the decision. If predictive analytics could help focus the campaign and generate the needed confidence, then you have an appropriate problem for your prototype to address.