In the attempt to contain the expansion of illegitimate products, a software regularization organization presented interest in knowing which of the people they intend to call will actually provide monetary results to the corporation. The company sought clients that had a genuine interest in leaving illegal software in the past and acquiring legitimate licenses. Furthermore, the organization also strived for a client large enough to present the company with profit that was worth calling. Such clients were extremely hard to find randomly in the marketing team’s calling list and demanded great effort and resources.
The issue was solved through machine learning models that optimized the work of the marketing team. The algorithm used Random Forests to, firstly, classify which of the clients in the calling list are using illegal software. Later on, we used a Regression algorithm to predict the amount of profit such clients would present to the company.
The graphic below highlights the importance of the features used in the Random Forest model. The importance of each feature is crucial for the understanding of the model and the data we are dealing with.
Notice that among the labels used, the number of PCs, age of the company, the number of branches, and the share capital seem to be the ones explaining the success or failure of the calls. Additionally, using such information, we were capable of building a model to analyze the data points and classify whether a certain client will buy licenses. For every 100 clients on the list, it is likely that 62 of them will purchase licenses after the call, which resulted in a gain of 92% in the process efficiency when compared to the previous precision.
Additionally, the regression model ranked clients from most to less likely to provide profit to the company, which was responsible for the optimization of the marketing team’s calls and better acquisition of capital for the institution.
At 4tune, we are passionate about solving clients’ problems to deliver profit. If you have a business idea or need help with anything related to artificial intelligence, we are glad to help. Please contact us at email@example.com for more details.