Student retention tips: which KPIs should you track?

Every year, education teams repeat the same cycle: a heavy push for new enrollments, then a quiet surprise when the classroom empties. The distance between those two moments is where most student retention tips stop being generic advice and become a measurement problem.

The Higher Education Map published by the Semesp Institute puts cumulative dropout for the 2020 to 2024 cohort at 64.7% in the private network, 68.1% in private distance learning and 69.2% at the largest institutions.

Those figures describe one entering group followed over years, not a single term. They say something simple to anyone working with educational marketing strategies: the enrollment link breaks slowly, and it leaves evidence long before the student disappears.

This article treats retention as a measurement discipline. What follows is a reading routine: which indicator to open, what each one shows, when to raise the alert and which campaign answers it.

What is student retention and where does it start?

Student retention is the work of keeping an active enrollment link until the diploma, not the single moment when a lead becomes a signed contract. It starts in recruitment, when expectation and reality are set, and it continues through every term in which the student decides again to stay.

Three words get used as if they meant the same thing, and they do not. Retention is an active enrollment link, re-enrollment is the renewal for the next period, and school dropout is the suspension or abandonment of the course.

Keeping them separate changes what you count. An institution that calls every non-renewal school dropout inflates the problem, and one that counts only formal withdrawals misses the students who simply stopped showing up.

Student recruitment and retention also start earlier than most teams assume. A promise made in the ad that the course does not deliver creates a student who is already leaving in the first term, so what the campaign promises and what the course delivers have to match.

Treating student recruitment and retention as one budget line also makes the trade-off visible. Every real spent filling a seat that empties in June is spent twice, once on the campaign and once on the replacement.

Acting while the student is still enrolled is cheaper than recovery. The routines that keep students in the classroom through the term are different from the ones built to bring someone back after the exit.

Caption: Why student retention starts as early as lead generation. Video in Brazilian Portuguese.

How does the education market look today, and why measure retention?

Knowing how the education market looks today is what turns retention from a vague worry into a budget decision. Volume, the split between modalities and exit rates define where the risk sits, and the Brazilian numbers describe a market that grew in enrollments while losing students at a pace no institution absorbs quietly.

The 2024 Higher Education Census released by Inep, the national institute for educational studies and research, records 10,226,873 undergraduate enrollments, with 5,189,391 in distance learning (50.7%), 5,037,482 on campus (49.3%) and 66.8% of new entrants in distance learning.

Distance learning, known in Brazil as EAD (Educação a Distância), is no longer a side channel. It is where most new students now enter, which means the retention routine has to work without a hallway, a coffee break or a professor noticing an empty chair.

Annual dropout tells the other half of the story. The Higher Education Map puts 2024 at 24.8% on campus, 26.6% in private programs and 21.4% in public ones, against 41.6% in distance learning, 41.9% private and 32.2% public.

The gap between modalities is the reason a single institutional average is useless. Reading how the education market looks today at that level of detail is what separates a report from a decision about where the next real goes.

This is also where educational marketing strategies stop being about reach alone. A campaign that fills a class the institution cannot hold costs twice, and the return on retention is what makes the second half of the budget defensible.

Entrance exam campaigns feel the same pressure. The promise made in the ad becomes the expectation measured in the first term, and a mismatch there shows up in the indicators long before anyone calls it a retention problem.

Which KPIs should you use to track student retention?

Five indicators cover most of what retention needs: satisfaction, perceived effort, attendance, failure rate and the dropout rate itself. Each one answers a different question, and the value comes from reading them together, because a student rarely leaves for a single reason that one number can capture.

What matters more than the list is the alert threshold. An indicator without a point at which someone has to act is a chart, and charts do not retain students.

Collecting the readings is no longer the hard part. AI (Artificial Intelligence) now handles the clustering and the ranking, so the scarce resource is the decision about what to do once the alert appears.

Net Promoter Score (NPS)

The NPS (Net Promoter Score) measures how likely a student is to recommend the institution, and it is the fastest read on whether the relationship is healthy. It is collected through a single rating question sent at defined moments of the term.

A falling score rarely means one bad class. It usually points at something structural, such as a service channel that does not answer or a schedule change nobody explained.

Open text answers matter as much as the score itself. The number says that something changed, while the comments say what to fix. This is where AI in student retention earns its place by turning thousands of open answers into a short list of themes.

Customer Effort Score (CES)

The CES (Customer Effort Score) measures how much work a student had to do to get something resolved, and it catches friction that satisfaction surveys miss. A student can like the course and still be exhausted by the institution around it.

Effort accumulates quietly. A request that bounces between WhatsApp, email and the front desk for a week teaches the student that the institution is hard to deal with, long before anything appears in a formal complaint.

Collect it right after a service interaction, not at the end of the term. The memory of the effort fades fast, and a late answer stops being useful for anything except a report.

Student attendance

Attendance is the earliest operational signal of disengagement, because absence usually starts before dissatisfaction is ever expressed. It is also the indicator most institutions already have, sitting unused inside the academic system.

Read it by subject, not only by student. When absence concentrates in one discipline, the problem is the discipline, and the answer is academic rather than a retention email.

In distance learning, attendance becomes access and activity. Logins, time inside the LMS (Learning Management System) and submitted assignments replace the physical roll call, and the drop is just as visible to anyone looking.

Failure rate

The failure rate shows where the course is losing students to difficulty rather than to disinterest, and it is the indicator that most often points at a cause someone can fix. Grades concentrated below the line in one subject are a curriculum signal.

Failure and school dropout are linked. A student who fails the same subject twice starts doing the arithmetic of time and money, which is why what pushes a student out of the course deserves attention before the withdrawal form arrives.

Student dropout rate

The dropout rate counts who left, and it is the only indicator on this list that reports a result instead of a risk. It closes the loop, because everything else predicts and this one confirms whether the prediction was read in time.

Compare it with the graduation and entry volumes of the same period. An institution losing more students than it graduates has a structural problem that no single campaign fixes on its own.

Each indicator becomes operational when it carries a threshold and an owner. This is how the five readings organize into a routine:

Table: How each retention indicator turns into an alert a team can act on.

The threshold is the part teams skip. Without it, the indicator gets reported every month and never acted on, and the dashboard becomes a record of exits instead of a tool that prevents them.

Caption: The five indicators behind the student retention tips, and the alert threshold each one triggers.

Why track by entry cohort and by modality?

Tracking by entry cohort and by modality is what makes a retention number comparable. A cohort follows the same group of students from the term they entered, and the modality cut separates on-campus from distance learning, where exit behavior is not the same.

Cumulative cohort dropout and annual dropout are different indicators, and they are not interchangeable. The first asks how much of an entering group had left by a given point, the second asks how many students left within a single year.

Mixing them produces conversations that go nowhere. A cumulative figure for a group followed over years will always look larger than an annual rate, and calling either one the dropout rate without saying which is being measured makes the comparison meaningless.

The cohort cut also tells you when students leave. Exits concentrated in the first term point at expectation and adaptation, while exits in the middle of the course point at cost, workload or a program that stopped making sense.

Modality changes the intervention, not only the number. A student who stops attending on campus can be approached by a coordinator in the hallway, and a distance learning student who stops logging in has to be reached by a system that noticed first.

That is the balance behind technology and human contact in recovery work: the alert is automated, and the conversation that follows is not.

Standardize the rule before comparing semesters. If one term marks a student as dropped out after a short window of inactivity and the next uses a longer one, the trend line is measuring the rule change instead of the students.

How do you turn the indicator into student retention campaigns?

An indicator turns into student retention campaigns when the alert triggers a specific message for a specific group. The reading defines the audience, the threshold defines the timing, and the channel defines the format, which is the opposite of sending one newsletter to everyone enrolled.

Email marketing strategies are the natural first channel because they carry context. A student flagged by attendance needs a different message than one flagged by failure rate, and email marketing strategies that ignore the difference read as noise.

Content marketing strategies work on the cause rather than the symptom. A group struggling with one subject responds better to study material and tutoring schedules than to a generic message asking whether everything is fine.

The second use of content marketing strategies is reputation. Students talk about the institution in public, and the channel where that conversation happens needs a measurement routine of its own, or a complaint becomes a permanent review.

Sales funnel strategies apply to retention more directly than it seems. The stages used to move a lead toward enrollment also describe the path from alert to conversation to renewal, and sales funnel strategies make that path measurable.

None of it survives without one record of the student. A CRM (Customer Relationship Management) platform holds the history, and a CRM set up for education is what lets an alert reach the right person with the previous conversation attached.

AI in student retention is most useful at the triage step. It ranks who is at risk and why, so a small team spends its hours on conversations instead of spreadsheets, and AI in student retention never replaces the conversation itself.

The segmentation built for entrance exam campaigns works here with almost no change. The lists are different but the logic is the same, which is why the plays that already worked at other institutions usually reuse infrastructure the marketing team already owns.

Educational marketing strategies aimed at enrolled students start from a list the institution already has. Digital marketing strategies for that audience cost less than the ones aimed at strangers, because the channel that attracted the student is often the channel the student still watches.

Marketing strategies for retention differ from acquisition in one respect: the audience is already known. Digital marketing strategies can therefore be precise about timing and message, which is a luxury acquisition never has, and marketing strategies built on that precision waste far less budget.

Student retention campaigns are built per indicator, not per calendar. An educational consulting partner usually earns its fee at this step, by defining thresholds the institution can honor with the team it actually has.

Caption: How to analyze student retention metrics and act on them. Video in Brazilian Portuguese.

Frequently asked questions about student retention tips

What should you prioritize first among student retention tips?

Among student retention tips, the first priority is not a new tool but a threshold. Pick one indicator the institution already collects, define the point at which someone has to act, name that person, and run it for a full term before adding the second.

Attendance is usually the right first pick. The data already exists, the signal arrives early, and the intervention costs far less than recovering a student who has already stopped paying.

The second priority is the cut. Reading the same indicator by entry cohort and by modality costs nothing extra and changes almost every conclusion drawn from the institutional average.

Only then does the campaign layer make sense. Campaigns built on top of a threshold nobody defined end up as generic email, and generic email is why so many teams conclude that retention work does not pay.

An educational marketing agency is useful precisely at the boundary between the number and the message. The institution knows its students, and an Educational marketing agency knows how to turn an alert into a sequence that gets answered.

The fastest way to retain students is to act on a number the institution already has. Everything else on this list is refinement of that first habit.

If you want to map which indicators your institution already collects and what is missing to act on them, talk to our team. We can review the current measurement routine and show where the first threshold should sit.