Migration

LMS migration checklist: what really moves, and what you rebuild

Illustration: a person checks a long clipboard next to open moving crates of booklets, certificates and folders
Illustration generated by AI.

Before switching platforms, answer one question: how much of the work is not automated? Some objects usually export cleanly, some almost always need rebuilding, and some depend entirely on both vendors' written answers. Active subscriptions are the classic blocker: they rarely move processors silently. The rule is simple. Migrate only once the unautomated work is quantified, not before.

The object matrix: three honest columns

List every object type you own, then sort it into one of three buckets. We do not claim any specific platform exports any specific object: export tooling is not in our registry, so the honest move is to inventory your own account and demand a sample export from each vendor during the trial. Use this as your inventory template.

How to inventory instead of guessing

Do not ask whether migration is possible. Count what you have, then measure what each vendor can actually deliver.

The reconciliation test

Prove the export before you trust it. Build a small fictional dataset on the platform you are leaving: five courses, twenty fictional learners with known progress, a couple of test orders, one issued certificate. Run the export, import it into the new platform, then compare object counts and a few field values before and after.

If twenty learners go in and eighteen come out, or completion dates arrive blank, you have found the gap while it costs nothing. Do this on fictional data first, never on your live school. The reconciliation check is the single most useful hour in any migration, because it replaces a vendor promise with a count you did yourself.

The go or no-go rule

Migrate only when the unautomated work is quantified. That means you can write a sentence like: courses and learners export cleanly, quizzes and drip rules need roughly thirty hours of rebuilding, and both vendors confirmed in writing that active subscriptions must be re-created by learners. That is a decision you can make. What you cannot decide on is a hopeful assumption that everything transfers.

Cutover order protects you: build and verify the new school completely, run the reconciliation test, then and only then point your domain and announce. Never touch DNS or tell learners before the destination is proven. If subscriptions are your revenue, treat their migration answer as the gate for the whole project.

Frequently asked questions

Can I move active subscriptions to a new platform?
Treat it as the hardest question, not a given. Active subscriptions usually cannot switch payment processors silently, because the card mandate stays with the old processor. Some vendors offer migration paths, others require learners to re-enter card details. Do not assume. Ask both vendors in writing exactly what happens to a live monthly plan, and get the answer before you commit to the move.
Does learner progress transfer between platforms?
Sometimes as a file, often not in a usable form. Progress and completion records are in the verify-with-both-vendors bucket for a reason. Request a sample export during the trial, open it, and confirm the completion dates and scores are actually present. If they arrive blank, you either rebuild the records or accept a fresh start. Decide that on real evidence, not a feature name.
How long should a migration take?
Long enough to run the reconciliation test first. There is no universal number, because the unautomated work varies with your quiz logic, community content and subscriptions. The honest sequence is: inventory, sample exports, a fictional-dataset reconciliation, then a rebuild estimate in hours. Once you can quantify the manual work, the timeline follows. Skipping the test is what makes migrations overrun.
Should I keep the old platform running during the switch?
Yes, until cutover is proven. Build and verify the new school in full, run reconciliation on fictional data, and keep the old platform live and billing normally until the destination passes. Only then point your domain and announce to learners. Cancelling the old subscription before the new one is proven is how schools lose access to records they still needed.

Keep deciding

Moving from Thinkific to LearnWorlds

Why teams move from Thinkific to LearnWorlds, SCORM without a quote and more admin seats, with plan mapping, cost math and a checklist-driven migration.

Moving from Teachable to Thinkific

Why creators leave Teachable for Thinkific, the 7.5% Starter fee, product caps, B2B selling, with plan mapping, cost math and a checklist migration.

Reporting and learner progress: what you can actually export

A training company must export per-client completion, scores and timestamps. What Thinkific and LearnDash publish, what LearnWorlds leaves to a demo.

Course platform pricing calculator: Thinkific vs LearnWorlds vs Teachable, Podia and LearnDash

Selling courses? Compare the yearly cost of Thinkific, LearnWorlds, Teachable, Podia and LearnDash from your sales, students, admin seats and transaction fees.