1. Introduction to Modeling


What is modelling, when do we use it and why is it so important? This short lesson will answer these questions and introduce our case study, Zippy Airways for this course

To explore more Kubicle data literacy subjects, please refer to our full library.


  1. What is a Model? (00:17)

    In Excel, a model is a simplified representation of reality that helps you make business decisions. A model simplifies the uncertainty of the real world and considers a more basic version of a problem. As a result, the specific answer produced by a model is generally wrong in reality. What’s more important are the insights the model provides about the business problem.

  2. When Should we Use Models? (01:06)

    In their book “Modeling for Insight”, Stephen Powell and Robert Batt identify badly structured problems as those lacking in some or all of the following characteristics:

    • Clear objectives for the analysis
    • Obvious assumptions
    • Necessary data is available
    • Well-understood logical structure behind the analysis

    If your business problem lacks some or all of these characteristics, then creating a model may help.

  3. Course Case Study (01:38)

    In this course, we create a model for an airline called Zippy Airways. They are considering whether or not they should implement an overbooking policy. Overbooking occurs when an airline sells more tickets for a flight than they have seats available. If some passengers don’t show up for their flight, then the airline generates extra revenue by overbooking the flight. However, if all passengers do show up, then the airline incurs a cost from moving some passengers to an alternative flight. We want to use a model to determine if an overbooking policy makes financial sense for Zippy Airways.


Modeling is a very powerful application of Excel. Most people have heard of modeling or financial modeling, but few ever learn how to build a fully functional model. If you're one of these people, then this course is for you. Let's get started by asking ourselves a question. What is a model? A model is just a simplified representation of reality that helps you make business decisions. So if I use models, well it turns out that most big business decisions are very difficult to solve. Laced with uncertainty and impossible to answer exactly. Faced with these decisions, we can rely on intuition or we can build a model. When building a model we create a simplified version of the problem that won't give us the exact answer but will provide insights. And that's the typical result from a model. Insights, not the correct answer. In fact, the only guarantee you will have from a model will be that the answer you find is probably wrong. But the insights along the way can vastly improve your decision-making process. So when should we use model? In their book, Modeling for Insight, Stephen Powell, and Robert Batt described ill-structured problems as those with few or none of the following four characteristics. Reading from this list, you'll notice that most decisions which require future forecasts are ill-structured problems and probably require a model. In this course, we're going to follow Powell and Batt's Four-Stage modeling process to help us make a business decision for a fictional company called, Zippy Airways.

Zippy Airways has been operating for three months out of Gatwick Airport in London and has 16 flights per day between London and four European cities. Brussels, Paris, Madrid and Berlin. Unlike most airlines, Zippy does not use an overbooking policy but the Chief Executive of the business has asked us to look at the flight data for the previous three months to see if the company should implement an overbooking policy on some or all of its flights. Overbooking is a policy airlines use to gain extra revenue from passengers who do not show up for their flight. Zippy operates Airbus A320s which each have a 180 seats. An overbooking policy would allow Zippy to sell more than this number of tickets on busy flights, say 190 tickets and hope that 10 people didn't show up for the flight. This would allow Zippy to collect more revenue per plane from the 10 extra tickets sold at no extra cost. However, if more than 180 people show up, then Zippy has to bump passengers to another flight which can be very expensive. In the last slide, I've included some additional detail on the case for you to review. Once you've done that, let's begin our modeling process in the next lesson with framing the problem.