Frequently asked questions
- What types of algorithms do you use?
Different Machine Learning algorithms are used depending on the solution. In all cases, the objective is to help institutions predict and solve problems related to student enrollment and retention.
- Do I need a technical team to implement this solution?
No! Our solution is plug & play. In other words, it is easily integrated with the systems of each institution (LMS, CRM, ERP, or SIS, depending on each case) and provides actionable information in real time.
- Do you have any proven success stories?
Yes! Universidad Siglo 21, the largest private university in Argentina. There, we implemented enrollment and retention solutions to improve enrollment and permanence. After the first year, it was possible to increase retention by 10.9 points (compared to estimates) and increase the number of students enrolled by 45%.
- Is it possible to adopt these solutions in public institutions?
Our solutions are developed to understand variables in private and fee-based institutions. However, it would be possible to analyze it under the dynamics of an institution that does not charge for its services.
- What are the most relevant variables?
Each solution has a different set of data and variables. For example, the student retention model focuses on:
- Personal and socio-demographic information that we obtain through SIS, CRM, LMS, and external sources.
- Student interaction on different platforms: chatbots, learning platforms, online seminars, and surveys, among others.
- Academic results (exam marks, for example).
- Student relationship with academic areas, and advisors, among others.
- Are the initial variables defined by Ed Machina?
It is a consensus between our experience and the needs of each institution. They can be 100% adapted to your particular variables, understanding that each institution, region, country, data source, and dynamics may be different.
5. Do the variables of the predictive model change according to the academic offer (undergraduate, undergraduate, graduate, etc.)?
The model learns from the data set we select to analyze. It is important to include different variables from the CRM, LMS, ERP, SIS; with the objective of enriching the model to learn and predict which is the variable that most affects each student, even according to the program, type of study, modality and cost.
6. How complex are the integrations with LMS, CRM, SIS or ERPs?
They don’t have any complexity. Our platform integrates through APIs to the most used systems in Education. Ed Machina’s technical team accompanies and advises the institution.
7. Integrations with LMS, SIS, CRM, how complex are they?
They have no complexity. It is a matter of knowing the systems with which the institution operates, analyzing the available data and how it is structured, and integrating it into our infrastructure. The information can be in the cloud or on-premise.
8. Can I integrate my lead management and marketing tools with student segments?
Yes. Smart Enroller ™ allows one to know the probability of lead conversion into a student with unknown variables that, otherwise, must be sought and studied outside. Thanks to Up and Cross Selling, we can know which product we can offer to a student.
Working on custom models generates clusters, with which student populations can be grouped and discover the characteristics of the leads that advance through the funnel for each institution. By identifying which lead is most likely to become a student, we provide a set of personalized recommendations on how to approach them for success.
- Is an internal team needed to implement the solution?
No, our solution is plug & play. Our technical team is available to provide support whenever the institution needs it.
- How do you ensure the privacy of the information?
For us, it is essential to ensure the privacy of the information of the institutions that choose us. In this sense, we develop a contract detailing all the points pertinent to privacy.
Once both parties agree with what is stated in this document, we are in a position to move forward with the implementation stage of our predictive models.