Media budget by program: how to split it without guessing

Few decisions in student recruitment weigh as much as splitting the budget between programs and campuses. It is made at the start of the cycle, when there is still little data for the period, and it usually follows what was done the previous term.

The result is familiar: programs that already sold well receive more investment and keep selling well, while classes that could have filled with a push are left without reach. The media budget becomes an inheritance rather than a decision.

There is a better path, and it does not depend on new tools. It depends on swapping the distribution criterion for one that considers how much each enrollment costs and how much it is worth in each program.

Why is distributing the media budget by cost per lead a mistake?

Distributing the media budget by cost per lead rewards the program that generates cheap contacts, not the one that fills the class. They are different things: a program can have the lowest cost per lead in the portfolio, the worst enrollment conversion rate and still receive the largest share of the budget.

Caption: the media budget stops being split by historical average and starts following the cost and value of each enrollment

Cost per lead measures an intermediate step. It says how much a completed form costs, which is useful for evaluating creative and landing pages, but insufficient for deciding investment.

The distortion appears in the full calculation. A program with a lead at ten units of currency and 2% conversion costs five hundred per enrollment, while another with a lead at thirty and 12% conversion costs two hundred and fifty.

Read by cost per lead, the first program looks three times better. Read by enrollment, it costs twice as much.

There is a second, less discussed reason. Programs with high spontaneous demand generate cheap leads because the audience already searches for them by name, so paid media simply harvests demand that would exist anyway.

Investing more in those programs inflates the campaign's apparent result. The report improves, and the number of new students at the institution does not change in the same proportion.

The third reason is the ceiling on places. No program absorbs budget indefinitely, and continuing to invest in a class that is already full wastes money that would make a difference elsewhere.

It is worth separating this discussion from a similar one. Understanding why cost per enrollment rises every term is a question of efficiency in student recruitment, while deciding how much each program receives is a question of allocation, and the two call for different answers.

How do you use cost and value per enrollment to allocate the budget?

The allocation method uses two numbers per program: the historical cost per enrollment and the value one enrollment generates while the student stays. The ratio between them produces an efficiency index, and it is that index, capped by the number of places, that orders the budget distribution.

Cost per enrollment comes from your own history. Add up the media investment attributed to the program in the previous cycle and divide it by the number of enrollments generated by that investment.

The value of an enrollment depends on three variables: average tuition, average length of stay and the program's margin. It is not the value of one term, but what the student generates until they leave or graduate.

Length of stay is the most forgotten variable and the one that changes the result the most. According to the 16th Map of Higher Education by Instituto Semesp, annual dropout in Brazil reaches 41.6% in distance learning, against 24.8% on campus.

With that difference, two programs with identical tuition can have very different enrollment values. Ignoring length of stay overestimates the return of the programs that lose the most students in the first year.

With both numbers in hand, the index is straightforward: value per enrollment divided by cost per enrollment. An index of 8 means every unit invested returns eight over the student's stay.

See how the ranking changes the reading in a hypothetical portfolio:

Table: illustrative example of ranking programs by the index between value and cost per enrollment, the basis of the media budget

In the example, program D has the cheapest lead and the cheapest enrollment in the portfolio and is the worst destination for additional budget. Program A is the most expensive per enrollment and the best return per unit invested.

Allocation follows the index up to the limit of available places. When program A fills its forty places, the budget moves down to the next on the list rather than continuing to push a program with no room for new students.

This method needs one piece of data many institutions still have not connected: the enrollment tied to the contact's source. Without that link in the CRM, cost per enrollment per program is an estimate, and the method loses precision.

How do you handle high-ticket, low-volume programs?

High-ticket programs with few places call for treatment separate from the main budget block. They can bear a high cost per enrollment, but do not absorb a high daily budget, because the audience is small and the campaign saturates before spending what was planned.

The practical rule is to reserve a dedicated budget, with a target in number of enrollments rather than lead volume. Ten enrollments in a medical degree or an executive postgraduate program can be worth more than three hundred in a low-ticket course.

The common mistake is applying those programs' cost per lead target to the rest of the portfolio. The manager sees the high cost, cuts the budget and loses the institution's most profitable enrollments.

These programs also have a longer decision cycle. Content, reputation and social proof weigh more than a discount offer, and paid media works better sustaining presence than forcing immediate conversion.

Each one's seasonality also weighs on the decision. Postgraduate programs usually have different entry windows from undergraduate ones, and concentrating all the budget on the undergraduate admissions calendar wastes the right moment for them.

Niche programs also benefit from qualified organic search. The work of local SEO applied to colleges serves the demand of someone looking for a specific specialisation in their own region.

How do you distribute the budget by campus and region?

Distribution by campus uses four variables: open places at the unit, demand potential of the location, local conversion history and competitive pressure in the region. The size of the unit does not enter the equation, because a large campus with a full class does not need additional budget.

The starting point is the number of open places per program at each unit. That is what defines the ceiling of useful investment for that location in the cycle.

Local demand comes next. Population in the relevant age range, the presence of secondary schools and search volume for the program's name in the region give the size of the available audience.

Video: the balance between paid media and organic that precedes any media budget decision, on the mkt4edu channel (video in Portuguese)

Conversion history distinguishes similar locations. Two cities of the same size can have very different enrollment rates, and the explanation usually lies in local service or in the unit's reputation.

Competition closes the reading. A location with three institutions competing for the same audience naturally has a higher cost per enrollment, and insisting on the same target as other locations leads to premature cuts.

There is an operational detail that often goes unnoticed. When radius campaigns overlap in neighbouring cities, the account competes for the same audience, and Google Ads states that, among eligible keywords from the same domain, only one can trigger an ad for the searched term. The overlap fragments budget without widening reach.

At institutions with many campuses, grouping similar locations into blocks solves much of the effort. Managing forty separate local campaigns takes more time than the gain in precision justifies, and blocks with similar behaviour simplify paid media management without losing control.

Communication also changes by location. A national campaign with a single asset pays off less than variations naming the city and the unit's address, especially in regions where the institution is still little known.

How do you review paid media allocation during the campaign?

Allocation review happens in weekly cycles, with small moves, and not in one big reallocation halfway through the campaign. Each week, the enrollment pace of each program is compared with the proportional target for the period and budget moves from those that already hit the ceiling to those running behind.

The review trigger needs to be defined beforehand. A rule such as "a program that reaches 80% of its places before the halfway point releases budget" avoids a discussion at every meeting.

Abrupt changes harm delivery. On Meta, significant edits send the ad set back into the learning phase, and Google advises waiting at least one conversion cycle before evaluating the effect of a change, so gradual adjustments preserve performance better.

The data supporting the review needs to arrive quickly. If the enrollment only shows up in the report twenty days later, the weekly decision leans on applications, which is the best leading indicator available.

Video: how to structure the measurement that supports media budget decisions, on the mkt4edu channel (video in Portuguese)

A measurement framework tied to revenue operations solves that delay, because it defines in advance which indicators answer for which decisions and at which frequency.

It pays to record every reallocation and its reason. At the end of the cycle, that history shows which decisions worked and feeds the initial allocation of the following period.

How do you build a media plan template by program?

A media plan by program is a spreadsheet with one row per program-and-campus combination, containing places, enrollment target, expected cost per enrollment, allocated budget and planned channels. The sum of the budgets closes the total, and each row supports its own target.

The minimum structure has eight columns. Program, campus, places, enrollment target, expected cost per enrollment, budget for the period, channels and owner cover what is needed to track without a parallel spreadsheet.

The enrollment target per row is not the same as places. It considers what recruitment needs to deliver after discounting re-enrollments, transfers and admission through other routes.

The expected cost per enrollment comes from history, adjusted for expected competition. When there is no reliable history, open with a conservative estimate and correct it in the second week.

Choosing channels per row avoids the trap of concentrating all paid traffic in one place. Programs with high search volume call for presence on the search network, while little-known programs depend more on discovery in social media to generate demand.

The plan needs to include a reserve. Setting aside 10% to 15% of the budget for reallocation during the campaign gives room to manoeuvre without requiring new approval at every adjustment.

Admissions exam campaigns concentrate predictable peaks, and sponsored links aimed at those periods need to enter the plan with their own budget, separate from continuous recruitment.

The plan should also talk to what happens outside media. In educational marketing strategies, in-person activity at schools and events feeds the same funnel, and a hybrid reading across online and offline channels avoids counting the same enrollment twice.

Frequently asked questions about the media budget

Where do you start redistributing the media budget?

Start by establishing the cost per enrollment of the previous cycle, program by program. Even an imperfect estimate changes the conversation, because it replaces a discussion about cost per lead with one about filling classes.

Next, estimate the value of an enrollment for each program using tuition, length of stay and margin. It is the half of the calculation almost nobody has ready and the one that most shifts the portfolio's order of priority.

After that, build the media plan with one row per program and campus, with a reserve for reallocation. From there, the weekly review does the fine tuning.

Teams already operating this way notice a change in the internal discussion. Student recruitment and retention stops being judged by the volume of leads generated and starts being judged by classes filled and students who stay.

That reading also changes the priority between programs. A program with a low cost per enrollment and poor student retention returns less than it appears on the spreadsheet, because the revenue disappears before the student completes the first year.

That is why it pays to cross allocation with retention data every term. Student recruitment and retention handled on the same dashboard avoids reinforcing budget in a program that fills the class and loses half of it along the way.

Good allocation depends on knowing how much each enrollment won is worth. The content on retaining and attracting students shows why a program that fills classes and loses students distorts any media plan.