How to Turn a YouTube Video Into Flashcards for Studying

Most people who try to make flashcards from a video end up with a deck of facts they copy-pasted from their notes, one bullet per card. It technically works, but it wastes the best thing a video-based deck can do: test whether you actually understood an explanation, not whether you can recite a sentence back. Here's how to build a deck that does the former.

Start from a transcript, not your memory of the video

Writing cards straight from memory after watching feels efficient, but you'll systematically under-cover anything you tuned out for, which is exactly the material you most needed a card for. Work from the transcript instead: read it once, and mark every place the speaker defines a term, makes a claim, or draws a comparison. Those three moments are almost always where a good card is hiding.

Write the question before the answer

It's tempting to lift a sentence straight from the transcript and turn it into an answer, then write whatever question fits. Flip the order. Decide what you actually want to be tested on, a definition, a cause-and-effect relationship, a comparison between two things the speaker mentioned, and write that as a real question first.

“What did the speaker say about X? vs. Why does X happen?”

Recognition vs. understanding, same source material

The first question tests recognition. The second tests understanding, and it's understanding that survives an exam.

• •One idea per card. A card testing three things at once is a card you'll rate "Hard" forever because you're never sure which part you missed.
• •Keep the source timestamp on the card, even if your app doesn't show it during review. When a card keeps failing, jump back to that second instead of rereading a static answer you've already half-memorized wrong.
• •Prefer a cloze deletion ("Spaced repetition works because ___ makes memory more durable than rereading") over open recall when the exact wording matters, like a formula or a named process.

Don't make a card for everything

A forty-five-minute lecture does not need eighty cards. If a fact is only useful in the context of the surrounding explanation, background color, a joke, an aside, it doesn't deserve its own card. A good rule of thumb: if you'd be comfortable being tested on it cold, out of context, in three weeks, it's a card. If you'd need the surrounding paragraph to make sense of the question, it's not ready yet; refine it or leave it out.

Review on a schedule, not all at once

A deck you make once and review once is barely better than notes you never reopen. The actual value of flashcards comes from spaced repetition: reviewing a card right before you'd naturally forget it, not the day after you made it and not a month later when it's already gone. If your tool doesn't schedule cards for you, a simple manual approximation is: review a new deck the next day, then three days later, then a week later, then roughly monthly, pushing back each interval only for cards you get right.