How a Head of School Evaluates an Online Curriculum Vendor

TL;DR

• Most vendor evaluations go wrong in the same place: the demo shows you the teacher's view, and you sign based on how the dashboard looks rather than on what a student actually does for the fifty hours that earn the credit.
• Four questions carry almost all the weight — what a student does in an hour, who is the teacher of record, what the per-seat math is against a hire, and what happens to your students' data. Everything else is a feature list.
• Ask for the student's hour first. A vendor that cannot show you an unedited one, on a live account, has told you something the brochure will not.

---

You have the meeting on the calendar. Someone on your board forwarded you a name, or a colleague at another school mentioned a platform, and now there is a forty-five minute Zoom where a very pleasant person is going to share their screen.

Here is what will happen. You will be shown a dashboard. It will be well designed. There will be a course catalog with more titles than you expected, a progress bar in a satisfying shade of green, and a slide near the end with three logos of schools you have heard of. You will leave the call thinking it looked good, which is the one impression the meeting was engineered to produce, and you will be no closer to knowing whether it works.

Vendor demos are built to be impressive. They are not built to answer the questions a head of school has to answer to a board, a parent, and eventually an accreditation team. Those are different questions, and you have to ask them on purpose, because the demo will not volunteer them.

What should I ask an online curriculum vendor before signing?

Four questions, in this order. The order matters, because the first one disqualifies faster than the other three combined.

1. Show me one student's hour, unedited. Not the teacher view. Not a highlight reel. One student, one lesson, start to finish, on a live account.
2. Who is the teacher of record, and what does the record say?
3. What does this cost per seat, against what the alternative costs?
4. What student data do you collect, why, and what happens to it when we leave?

Everything else — integrations, reporting, mobile support, the roadmap — is real but secondary. If those four have good answers, the feature list will sort itself out in the pilot. If they do not, no feature list rescues it.

Ask to see what a student actually does

This is the question that separates a curriculum from a content library, and almost nobody asks it, because the demo is structured to keep you in the administrator's seat.

Ask to sit in the student's seat instead. Then watch for what the hour is actually made of. In a great many products, the honest answer is: a video, then a multiple-choice quiz, then a video. That is not a course. That is a playlist with a gradebook attached, and students treat it exactly the way you would expect — they scrub the video, they guess the quiz, and the progress bar turns green while nothing happens.

The research on this is not gentle. In a multisite randomized study of online credit recovery, only 28% of Algebra 1 students and 34% of English 9 students completed at least three-quarters of their online course, with the average student finishing 51% of the content. Half a course, on average. That is what "self-paced online curriculum" means in practice at a lot of schools, and it is why the veteran teacher on your faculty gets visibly tired when you mention buying one.

So make the vendor show you the hour, and ask three follow-ups while you watch:

• Where does the student have to produce something? A sentence they wrote, a decision they defended, a project an adult can look at. If the entire hour is consumption and selection, the credit will be hard to defend later.
• What happens when the student gives a lazy answer? Type "idk" into whatever the platform offers and see what it does. This is the single most informative thirty seconds of any evaluation. A tutor that accepts "idk" and moves on is a tutor that will accept it four hundred times this semester.
• What is left at the end? Ask to see a finished student artifact — not a score, a piece of work. If the platform cannot produce one, then at the end of a semester you will have a percentage and nothing to show a parent.

That last one is worth being stubborn about. When we built our own courses, the thing we cared most about was what sits at the end of the pathway, which is why every one of them terminates in a project rather than a final exam. You can look at those directly on the courses page rather than take my word for it.

Who is the teacher of record?

Ask it plainly, and do not accept a product answer to a compliance question.

Some vendors will tell you they provide the certified teacher. That can be genuinely fine, and it changes your cost structure, your scheduling, and your accreditation paperwork in ways you need to see in writing before you sign. Some will tell you the platform handles instruction — which means, when the accreditation team asks who taught the course, the answer had better be a person on your staff with a name and a credential.

The model that holds up best at small schools is the one where your teacher stays the teacher of record and supervises the room while the platform carries the instruction. That keeps the credit yours, keeps the transcript defensible, and keeps the person who knows the student in charge of the grade. It is worked through in detail in teacher of record for AI-delivered courses, and it is the question your accreditor will ask in some form.

While you are there, ask the seat-time question in the same breath, because the two travel together: how does this vendor document instructional hours, and does that documentation match what your accrediting body actually requires? The answer is often more flexible than heads of school assume, and it is worth knowing the rules before you let a vendor tell you what they are — the specifics are in do online electives count toward accreditation seat time.

What does it actually cost per seat?

Vendors quote a total. Boards approve a total. Neither number tells you whether the decision was good.

Do the math the other way. The relevant comparison is not "this platform versus that platform." It is "this platform versus the thing we would otherwise do," and the thing you would otherwise do is either hire someone or not offer the course. A teaching line at a small private school runs $47,000 to $70,000 a year all-in. That is the number on the other side of the ledger, and it is why a per-seat program can add four electives for less than a fraction of one hire.

For reference, our own school pricing is $149 per student per year, $129 at 25 or more students, and $109 at 50 or more, with a Founding School rate of $99 per student for year one for the first ten schools. Middle school Explorer seats are $49. Those are on the pricing page and I would rather you compare them against a competitor's real numbers than against a vibe.

Three things to insist on while you are pricing:

• Get the all-in number. Onboarding fees, per-teacher licenses, minimum seat counts, and the year-two rate. The year-two rate especially. A first-year discount that resets to a number you cannot afford is a two-year problem disguised as a one-year win.
• Ask what happens to seats you do not use. If you buy 40 and enroll 28, what happens? Some vendors will let you reallocate mid-year; some will not.
• Price it against the hire, in writing, on one page. Your board does not need a product comparison. It needs to see what a course costs three ways: hire, vendor, don't offer it. The full version of that exercise is in private school elective budget planning.

What about student data and privacy?

Short section, high stakes, and the one place where a confident-sounding answer is most likely to be wrong.

The core questions are the same ones any technology leader would ask: what student data the vendor receives, why it is necessary, how it is protected, who else can access it, and what happens to it when the contract ends. Vendors should be requesting the minimum personally identifiable information needed to do the job, and you should ask directly whether any student data is mined, and for what — mining for advertising directed at students or families is a bright line.

One useful piece of skepticism: "FERPA compliant" is not a certification. There is no certifying body. When a vendor says it, the right response is "what do you mean by that," and the quality of the answer tells you a great deal about who you are dealing with.

Then ask the exit question, which almost nobody asks during a sales cycle: when we leave, what do we get back and what do you delete? You want student records, grades, and completed work exportable in a form you can keep, and a deletion commitment with a timeline. A vendor that has thought carefully about how you leave is usually a vendor worth staying with.

How do I run the pilot so it tells me something?

Small, real, and measured — and decided in advance.

Do not pilot with a hand-picked group of your most motivated juniors, because you will learn nothing except that motivated juniors finish things. Pick a normal cohort in a normal period with the teacher who is most likely to be skeptical, and write down before you start what success looks like: how many students complete, what the finished work looks like, and how many hours a week the supervising teacher actually spends.

That last number is the one that surprises people, in both directions. Ask the vendor to predict it, then measure it. A platform that quietly costs your teacher eight hours a week is not saving you a hire.

Two structural notes. First, run it long enough for the novelty to wear off — four to six weeks minimum, ideally a full quarter. Enthusiasm in week one is not a signal. Second, decide the go/no-go criteria before the pilot starts, in writing, or you will end up relitigating the decision on vibes in a faculty meeting in November. The sequencing for all of this is laid out in how to run an elective pilot at a small private school.

If you are evaluating for middle school rather than high school, the pilot shape is different and shorter — exploration rotations rather than semester courses — and the middle school wheel is built for exactly that.

The honest summary

The evaluation that works is not longer than the one you were going to do. It is just pointed somewhere else.

Stop evaluating the dashboard. The dashboard is the part the vendor spent the most money making impressive, and it is the part you will look at least. Evaluate the student's hour, because that is where the credit is either earned or quietly not earned, and it is the only part of the product your students will experience at all.

Ask for the unedited hour first. Type "idk" and watch what happens. Then ask who signs the transcript, what a seat costs against a hire, and what happens to your students' data when you leave. Four questions, one honest hour of watching, and a small pilot with the criteria written down in advance.

If a vendor makes that easy, you have learned something good about them. If a vendor makes it hard, you have learned something more useful.

---

Start Your Pilot

If you want to run this evaluation on us, start where I would start: ask to see a student's hour. Start a pilot or book a school demo and we will put you in a student account, not a sales deck. School pricing is on the pricing page, and the full catalog — 31 high school courses across six career pathways, plus 24 middle school Explorer courses — is at /courses.

---

About the author

Steve Smith is the founder of Elective Genius and spent 25 years in education before building it. He also runs Rosedale Christian Academy, where every middle- and high-school student uses the platform — which means every decision here gets tested on his own students first. Reach him at schools@electivegenius.com.