In some instances, you need not have only one model to predict the outcome. If you accept the view that All models are useful, why not use some of them at the same time? This is used extensively in marketing campaign management and financial portfolio management. Sometimes it pays not to put all of your eggs in one basket. One model can fail to work during one month but is offset with positive results from another model. Ensemble modeling, which is similar, can also incorporate several diverse models at the same time, but usually there is a single outcome and may be subject to overfitting.