Conversion forecasting for E-Learning Platform

E-learning platforms have repeatedly proven their effectiveness, and their popularity has grown several times over in recent years. For example, according to Global Industry Analysts, the E-learning market will increase eightfold in just eight years (from 2022 to 2030). Statista predicts that by 2027 there will be 1 billion online students worldwide! These figures are inspiring; however, even well-known E-learning platforms face the challenge of retaining students at the early stages of interaction. We know exactly how to address this problem, and we are sharing one of the solutions with you.

Customer profile

Our client provides distance learning services for students who want to learn about marketing, SEO, and copywriting from scratch or expand their existing knowledge. They have already served over 15,000 students, who have successfully completed their courses and received certificates through the E-Learning platform.

Initial Challenge

The company communicated with potential students through a call center. This is an expensive communication channel, but it is not advisable to abandon it due to its effectiveness. Therefore, there is a need to pre-evaluate the leads in order to optimize communication and to call only those leads who are highly likely to start their studies. Ultimately, the company strived to increase the proportion of students who would commence their studies after registering and to optimize communication costs.

Our Solution

Developing a forecasting model that allows identifying those students who are highly likely to purchase the educational course after registration without any additional communication. The following tasks have been completed:

• Collected and analyzed the available data for modeling, assessing their potential use for the client’s needs.
• Identified over 500 metrics that may influence the completion of a purchase — that is, the final decision to undertake a course, confirmed by signing a contract and making a payment.
• Deployed an AI engine in the cloud provider’s infrastructure, which provides computational resources for the E-Learning platform.
• Developed a probability purchase forecasting model based on historical data accumulated over the platform’s operation.
• Conducted diagnostics of the factors influencing the likelihood of making a purchase.

We determined which factors most significantly influence the decision to purchase and proposed ways to leverage them to address the client’s goals.

Outcome

Our client used the information obtained to change their approach to communicating with potential students.

The tactics chosen by the E-Learning platform included:

• Users with a high likelihood of buying have been switched to email communication. Call centers are only used for “closing” when necessary.
• Users with a low likelihood of buying only received the email chain, saving a significant amount of communication budget on non-perspective leads.
• Users who were hesitant and couldn’t make a final purchase decision were immediately called by a manager.