Then a friend asked me what the best episode I'd heard that month was about. I named the show and the guest. I could not tell her a single idea from it.

That's the trap with audio. It's the easiest way to take in information and one of the worst ways to keep it. You're usually doing something else with your hands and eyes, there's nothing on screen to anchor the idea, and episodes run long enough that the good part from minute 22 is gone by minute 50.

Listening less won't fix it. Neither will sitting at 1x with a notebook open; I tried that and it lasted about a week. What finally worked for me is a small set of habits that fit around the way people actually listen.

Why podcast memory is so bad

There's a reason the forgetting feels worse than with books. When you read, you control the pace. You can stop on a sentence, reread it, underline it. A podcast keeps going whether you've understood the last point or not, and you rarely rewind because you're driving or your hands are wet.

You also get almost no retrieval practice. Hermann Ebbinghaus showed back in the 1880s that memory drops off steeply in the first day or two after learning something, and that each successful recall slows the drop. Most of us hear an idea once, nod along, and never try to pull it back up again. So it fades on schedule.

Speed doesn't help much either. I listen at 1.5x, and I'll admit that at 2x my retention fell off a cliff. If you're struggling, try knocking the speed down for shows you actually want to learn from and keep the fast setting for the ones you're listening to for fun.

Mark the moment, not the episode

Saving a whole episode is nearly useless. "That long interview with the sleep researcher" might be three hours of audio. You're never going back to scrub through it.

What you want is the timestamp. Many podcast apps, Spotify included, let you share an episode from the current position. When something lands, grab that link before you lose it. If you can't, say "remember 34 minutes" out loud and write it down the next time you're stopped.

One line of context makes a big difference. "34:10, why she says caffeine after 2pm wrecks deep sleep even if you fall asleep fine." Future you will thank present you.

Say it back within an hour

This is the cheapest habit on the list and the one that helped me most. After an episode that had something good in it, explain the main idea out loud in about thirty seconds. To your partner, to a coworker, or to nobody in the car.

If you can't do it, you didn't really get it, and now you know where to rewind. If you can, you've just done your first recall, which is the thing that slows forgetting down.

Turn ideas into questions

Notes like "sleep stuff, interesting" are what I used to write. They're worthless a week later.

Instead, write one or two questions the episode answered. "What does adenosine have to do with feeling sleepy?" "Why is a consistent wake time more useful than a consistent bedtime?" A question forces you to pull the answer out of memory later, and that act of pulling is what makes it stick. Rereading a summary feels productive, but mostly it builds the feeling of knowing.

Bring it back on a schedule

Here's where almost everyone falls off, me included for a long time. You need to see those questions again: the next day, a few days later, about a week after that, then a few weeks on. Each review takes seconds. If you remember the answer, push the next one further out. If you don't, bring it back sooner.

That's spaced repetition. Language learners have used it for decades with flashcard apps like Anki, and it works just as well for a podcast about negotiation or the history of Rome. You don't need many cards. Two per good episode adds up fast over a year of listening.

Use one idea that week

The ideas I remember best from podcasts are the ones I acted on. I moved my coffee earlier after that sleep episode, and I can still tell you why. I heard a dozen productivity episodes that year that I can't recall at all, because I never changed anything.

Pick one idea per week and try it. Small is fine.

Where I ended up

I did all of this by hand for a while with a notes app and calendar reminders. The thinking was the useful part. The clerical work was where I kept dropping it: copying links, finding the timestamp again, remembering to review.

Recl brings key ideas from your saves back on a schedule.

So I built Recl to handle that side of it. You save a podcast, video, reel or article in one tap and it writes a summary with the key ideas and timestamps, so the minute-34 moment is right there. It files the save into a topic on its own and then brings the key ideas back on a spaced-repetition schedule. If you only half-remember something you heard, you can search your saves by meaning or ask a question, and the answer links back to the source. It's on Android, web and Chrome, with iPhone coming soon.

A notebook and a calendar will get you most of the way, though. Tonight, pick the best episode you heard this week and try to explain its main point in thirty seconds. If you can't, that's your answer.

About the author

Spartan Rich is the maker of Recl, an AI learning app that helps you remember the videos, podcasts and articles you save, with timestamped summaries, automatic topics and spaced repetition.

Open your Watch Later list. Scroll to the bottom.

There's probably a conference talk you were excited about two years ago. A tutorial for a tool you've since stopped using. A podcast episode someone recommended in a meeting. Each one was saved with good intentions, and each one quietly expired.

Saving feels like progress. It's a tiny promise to your future self: I'll get to this. But a list with no process behind it doesn't teach you anything. It just moves the guilt somewhere tidier.

More discipline won't fix it, and neither will a strict "watch everything" rule. What helped me was a simple system that decides what gets learned, brings it back, and lets the rest go. Mine has five parts, and none of them takes long.

Part 1: Capture with a reason

Most saves fail at the moment of saving, because they carry no context. Three weeks later you look at a thumbnail and can't remember why it mattered.

One-tap save: Recl reads the video and pulls out the key ideas in the background.

So when you save something, add one line saying what you want from it. "How they structure code reviews for a small team." "The bit about pricing annual plans." "Need this for next month's migration."

If you can't write the line, that tells you something too. It probably wasn't worth saving.

Keep everything in one place as well. If learning-worthy content is spread across YouTube, a podcast app, Instagram saves and browser bookmarks, what you have is five graveyards.

Part 2: Triage weekly

Pick one slot a week, maybe 20 minutes on a Friday afternoon, and go through what you saved. Each item goes into one of three piles.

Learn is for anything that connects to something you're working on or want to get better at. You'll process those. Reference is for things worth finding later but not worth memorizing, so tag them and move on. Delete is for things that were interesting at the time and are irrelevant now. Let them go without guilt.

The delete pile is where this system earns its keep. A smaller list is one you'll actually deal with.

Part 3: Turn content into questions

For anything in the Learn pile, skip the summary. Write two or three questions you should be able to answer once you've understood it. For a 45-minute talk on incident response, that might be "What should the first message in an incident channel include?" and "Why does the speaker separate the incident lead from the person fixing the problem?"

Next to each question, note a timestamp or section heading where the answer lives. If you blank on it later, you can jump straight to the source instead of rewatching from the start.

Questions make you retrieve an idea. A summary only lets you recognize it, and recognition feels like knowledge without building much of it.

Part 4: Schedule the comeback

Almost everyone skips this part. It's the one that matters most.

The Recl feed brings key ideas from your saves back to you.

Memory fades fast after you first take something in, and more slowly each time you successfully recall it. Spaced repetition uses that by bringing each idea back at growing intervals, right around when you'd otherwise forget it.

A practical schedule is the next day, then 3 days later, then about a week later, then a few weeks after that, and finally a couple of months on. Each review is quick: read the question, answer it in your head or out loud, then check. If you got it, push the next review further out. If you missed it, bring it back sooner.

A few minutes a day covers a surprising amount of material.

Part 5: Use it within a week

Applying something is the fastest way to lock it in. After processing an item, look for one small way to use it in the next seven days. Try the technique in your next pull request, or bring the idea up in a one-on-one. Writing a short post that explains it in your own words counts too.

You don't need to apply everything. One real use a week is enough to change how much sticks.

The sticky-note version

If five parts sounds like a lot, this fits on a sticky note:

Save with a reason. Triage on Fridays. Write two questions. Review at growing gaps. Use one thing a week.

Automating the boring parts

I ran this system by hand for a long time with a notes app and calendar reminders. The thinking was worth doing. The clerical work (copying links, finding timestamps, remembering when to review) was where I kept falling off.

Ask: questions answered from your own saves, with links back to the sources.

That's why I built Recl. You save videos, reels, podcasts and articles in one tap. It summarizes them with timestamps, sorts them into topics automatically, and uses spaced repetition to bring key ideas back before you forget. When you remember saving something but not where, semantic search or Ask will look across everything you've saved, and the knowledge graph shows how your topics connect.

A notebook works fine too. Whichever you use, try one Friday triage this week and see how much of your Watch Later list survives it.

About the author

Spartan Rich is the maker of Recl, an AI learning app that helps you remember everything you save, with timestamped summaries, automatic topics and spaced repetition. Recl is freemium and available on Android, web and Chrome, with iPhone coming soon.