Provide ABM with historical data about your clients and their behavior. This will be the basis for ABM to learn and build predictive models. The scope of information may include: age, income, place of residence, results of previous campaigns, etc.

You will be asked a few questions which will help ABM to understand your goals and preferences. For instance: what do you want to predict, how fast should the whole process be, are there any characteristics you want to exclude from the analysis, what are the expected profits and costs regarding the campaigns you want to conduct?

Check out the reports with comprehensive information on the model. These reports exist to show the outcomes and quality of a model, to explain what data is used in the model, and what influence the data has on client behavior. You can also see how much profit you might gain from the campaign based on ABM predictions; make changes in the model parameters if necessary.

Now let ABM apply the model to new data and generate predictions. As a result, you’ll receive a .csv file with a list of you clients and predicted target value assigned to each of them. You can also download a scoring code and use it in your custom process.