What a Private School Network Loses When Each Campus Runs Admissions Its Own Way

Six campuses, six yield rates, no answer

You run a network of private schools, and the enrollment report lands on your desk every spring with the same problem. One campus converted most of the families it admitted. Another converted well under half. The board wants to know why, and the honest answer is that nobody in the building can tell them.

Maybe the weak campus has a demand problem: wrong neighborhood, wrong price, a competitor that opened down the road. Maybe it has a process problem: slow decisions, a visit day that does not sell, financial aid letters that arrive after families have already committed elsewhere. Those two diagnoses lead to completely different spending decisions. One says market harder. The other says fix the office.

The report cannot tell you which one you have, and it is not because the data is missing. It is because the six numbers on the page are not measuring the same thing.

The loss is comparability, not efficiency

When people describe the cost of campus-by-campus admissions, they usually describe duplication. Six admissions directors, six inquiry forms, six spreadsheets, six subscriptions to slightly different tools. That cost is real, but it is small, and it is the argument vendors make because it is the argument their product answers.

The bigger loss is that the network cannot compare itself. Yield is admitted families who enroll, divided by families admitted. Simple, until you ask each campus what "admitted" means.

At one campus, a family is admitted when the committee votes. At another, it is when the offer letter goes out, which might be two weeks later. A third counts waitlist offers that convert in August. A fourth admits siblings of current students outside the cycle entirely and never records them as applicants. Run yield across those four and you get four numbers that share a name and nothing else.

The same is true one stage up. Ask six campuses what a "complete application" is and you will hear six answers. Does it need the teacher recommendation? The assessment? The visit? The financial aid form, for families applying for aid? A campus with a loose definition will show a fat applicant pool and a thin acceptance rate. A campus with a strict one will look selective and small. Same families, same demand, different paperwork.

Melt and time-to-decision have the same problem. Melt depends on when you stop counting. Time-to-decision depends on when the clock starts. If each campus picked its own boundaries, and in a network that grew by acquisition or by letting strong heads run their own shop, each campus did, the network is reading six dialects and calling them a language.

This is not a peculiarity of schools. We see the identical pattern in franchise groups that try to roll up unit reporting, where the hard part is never the rollup, it is getting everyone to agree on what labor percentage means.

Why the sector benchmarks do not rescue you

The natural move is to check your campuses against the outside world. The sector does publish numbers. NBOA's five-year trend report, covering 279 day and boarding schools from 2020-21 through 2024-25, found median application coverage (applications divided by new students) rose from 242% to 262%, and that 60% of schools improved on that measure over four years.

That is useful context and a poor diagnostic. A school's position against a national median tells you nothing if the school's own count of "applications" follows a different rule from the next campus over. Benchmarks assume the definitions are settled. Inside your network, they are not. You cannot benchmark your way out of a definitional problem. You can only benchmark it into a more expensive-looking chart.

What vendors sell instead

Search for help with this and page one is admissions software: Blackbaud, Rediker, Enrollsy, DreamClass, and others. Each sells network-wide workflow configuration. One system, every campus, identical forms and checklists.

None of them separates two different things, which is the whole point of this article:

1. Standardizing definitions. What counts as an inquiry, an application, a complete application, an admit, an enrolled student, a melt.
2. Standardizing process. When applications open, which assessments are required, how many visits, who sits on the committee, when decisions go out.

Software makes the second one easy to impose, so it gets sold as the answer to the first. It is not. You can put six campuses on one platform and still have six definitions of "complete," expressed as six differently configured checklists. A migration like that produces the same incomparable yield numbers, now in a nicer dashboard.

Where centralization projects actually die

The opposite instinct, centralize everything, fails in a predictable place.

Campuses in a network rarely face identical conditions. School Pathways, which sells enrollment software and so has every reason to argue for centralization, concedes the case against it: campuses operating under different authorizers or in different states face different regulations, application timelines, and eligibility rules, and a single workflow forces a bad fit on some of them.

Even within one state, the timeline is local. A campus competing with a selective day school that releases decisions in March cannot run a rolling admissions calendar that suits a campus in a market where families decide in June. A campus with a boarding program needs interview logistics that a day campus does not. A campus serving a particular faith community may have requirements no other campus in the network shares.

Centralization projects that try to flatten those differences produce a predictable sequence. The network issues one calendar. The campus heads with the strongest markets push back, because the calendar costs them families. Exceptions are granted. The exceptions become the rule. Two years later the network has a central admissions function, a shared platform, and the same local variation it started with, plus resentment.

That failure is what makes a lot of network leaders give up on the whole idea and leave campuses alone. Which brings them back to the spring report they cannot read.

The split that works

The answer sits between those two failures, and it is narrower than either.

Standardize the definitions and the stage gates. Every campus counts an inquiry, an application, a complete application, an admit, and an enrollment by the same rule, and records the date each family crosses each line. That is a governance decision, not a software purchase, and it costs very little to make.

Do not standardize the timeline or the local requirements. Let the campus that needs an early decision date have one. Let the boarding campus run its own interview process. Let local admissions directors do the selling, which is what they are good at and what a central office is usually bad at.

With shared definitions, yield at campus A and yield at campus B finally mean the same thing. A gap between them is now a signal. And because every campus records when families cross each gate, you can see where in the funnel the gap opens. A campus that loses families between admit and enrollment has a different problem from one that loses them between inquiry and application, and neither problem is solved by the other's fix.

This is the comparability test, and it is worth running before anyone signs a software contract. Pick one metric, say yield. Ask each admissions director to write down, in one sentence, exactly which families are in the numerator and denominator, and on which date they get counted. If the sentences do not match, the network's enrollment report is not a comparison, and no platform will make it one until they do. We have no public dataset on how widely yield varies across private school networks, and we would be suspicious of anyone who claims one, precisely because the definitions underneath it are rarely aligned.

It is the same root cause we see in student information systems that go wrong in the same three places: records break where a definition changes hands, not where the software is weak. And as with district enrollment paperwork, the most useful number is often the one nobody has agreed how to count.

When software does belong

None of this is an argument against admissions platforms. Once the definitions are agreed, a shared system is the cheapest way to enforce them, because the stage gates become fields rather than habits. A network that standardizes definitions first will buy a smaller configuration, migrate faster, and get a report it can actually read on day one.

A network that buys first will spend the implementation arguing about definitions anyway, only now with a consultant from the vendor billing for the meeting. Which is a fine way to learn the difference between a stage gate and a toll gate.

The sequence matters more than the tool. Definitions, then gates, then software, and the timelines stay local throughout.

The question for your network

If your enrollment report shows campuses that look very different and nobody can say whether the difference is demand or process, the problem is probably not your admissions teams and probably not your software. It is that the numbers were never built to be compared.

Working out which definitions your campuses actually use, which ones need to converge, and which local differences are worth protecting is a week of conversations with the people who run admissions, not a procurement exercise. If that is the conversation your network needs, talk to us before the next enrollment cycle locks in another year of numbers nobody can read.