Last month I cut the same 47 second clip three times for one show. Same words, same laugh at the end, same reframe. One version went out on YouTube Shorts, one on Reels, one on TikTok. Three different edits, three very different outcomes, and the host asked me the question every host eventually asks: which one is actually worth the effort?
Here is my answer after two years of watching these numbers land. They are not three copies of the same channel. They are three different machines, and the one that matters most for a podcast is almost never the one hosts assume.
I am going to be blunt about it, because the polite version of this advice ("post everywhere, repurpose everything") is how people burn six months producing content that goes nowhere.
The three platforms want three different things
TikTok wants a conversation you were not supposed to hear. It rewards specificity, tangents, opinions with a body count. The algorithm there is still the most willing to show a clip from an account with 300 followers to 40,000 people, which is why it stays the best discovery surface for a show nobody has heard of yet.
Reels wants a moment. Short, punchy, visually legible on a phone held at arm's length in bad light. Instagram's audience is browsing between things. A clip that needs 20 seconds of setup dies there. A clip that lands its point in eight seconds does fine.
Shorts wants a trailer. This is the part people miss. YouTube Shorts is the only one of the three that sits inside the same account as your full episode. Someone watching a Short is one tap from your channel page, where the 58 minute version is waiting. Nothing on TikTok connects to anything.
So the question is not "which platform gets more views". The question is what you want the views to do.
What I actually tell hosts to do
If your show is under a year old and you want people to discover it, TikTok first. If your show already has an audience and you want them watching full episodes, Shorts first. Reels third, always, unless your audience is specifically on Instagram for professional reasons.
That ordering upsets people. Reels feels like it should matter more because Instagram is where everyone already is. But Instagram is where everyone already is with their existing friends. It is a maintenance platform, not a growth platform, and treating it like a growth platform is how you end up with 4,000 followers who have never listened to an episode.
The clip that gets the most views is rarely the clip that gets the most listeners. I would rather cut one Short that sends 200 people to a full episode than one TikTok that gets 90,000 views from people who will never learn the show's name.
The edit is not the same edit
This is the practical part, and it is where most repurposing tools fall over. They export one vertical file and blast it to all three. Then hosts wonder why performance is uneven.
Here is what actually changes between the three, as I cut them:
- Caption size and placement. TikTok's UI eats the bottom third with the caption, username, and music strip. Reels eats a similar band but the safe zone is slightly different, and Instagram will crop the top if you are careless. Shorts is the most forgiving. I keep captions in the middle band for TikTok and Reels, and let them sit lower on Shorts.
- The first frame. On TikTok, motion in frame one buys you the second one. On Shorts, the thumbnail frame matters because Shorts sometimes surfaces as a still. I pick a frame where the host's face is open and the mouth is not mid-word.
- Length. TikTok tolerates 60 to 90 seconds if the story is worth it. Reels wants 15 to 30. Shorts sits in between, and I keep them under 55 seconds so they never risk falling out of the Shorts feed.
- The ending. TikTok clips can just end. Shorts should hand off. A one line "the full conversation is on the channel" costs you two seconds and does more work than any hashtag ever has.
That is four decisions, not four exports. It is also why we price vertical clips as an add-on at $25 each rather than pretending they fall out of the main edit for free. They do not. Someone has to watch the episode, find the moment, and cut it three ways.
The mistake I see most is treating clips as something you find afterwards. The best ones are not found in the edit. They are made in the room. A host who knows they will need six verticals will naturally restate the question before answering, pause after a punchline, and avoid burying the good line inside a 90 second ramble that cannot be extracted.
I have cut episodes where I watched the whole thing twice and found two usable moments. I have cut others from hosts who think in clips and pulled twelve without trying. The difference is not talent, it is intent.
If you are struggling to get clips out of your episodes, the problem is usually upstream, and it often connects to how long your episodes actually run. Long is fine. Shapeless is not.
Volume is a trap, but so is perfectionism
The advice going around is to post three to five times a day per platform. For a podcast, that is nonsense. You do not have that much good material, and posting weak clips trains the algorithm to show your work to people who scroll past.
But the opposite failure is worse. Hosts who wait until they have the perfect clip post twice a month and never build any signal at all. My honest recommendation is four to six clips a week, distributed across platforms rather than duplicated across all three every time. Not every clip needs to appear everywhere. I have written more about finding a rhythm you can actually hold, because the sustainable number beats the theoretically optimal number every single time.
Along the same lines, this is really a question about whether the whole operation survives contact with a normal week. A show that dies in month five was almost always over-designed in month one, which is a thing I keep coming back to when hosts ask me how to build something that survives their actual schedule.
Guests change the math completely
Here is where the platform ranking flips. If your show runs guests, the single highest leverage clip is not the one that performs best. It is the one the guest will actually repost.
That means the guest looks good, is framed well, says something they are proud of, and gets tagged properly. A guest with 40,000 followers reposting one Reel is worth more than three TikToks that do nothing, and Instagram is usually where guests repost, because that is where their professional identity lives.
So Reels moves from third place to first place the moment guests are involved. Not for your audience. For theirs. Most shows never build a system around this and just hope the guest shares something, which is the same as planning nothing. I wrote a whole piece on making guest promotion something that reliably happens rather than something you wish for.
What none of this fixes
Clips do not fix a show nobody wants to hear. I want to be clear about that, because there is a whole industry selling the idea that distribution is the bottleneck.
Sometimes it is. More often the episode is fine and the packaging is not. A clip that goes out with a lazy caption, a bad first frame, and no channel handoff is not a distribution problem, it is a craft problem.
The same is true one layer up. If your Shorts are sending people to your channel and your full episodes have thumbnails that look like a screenshot, the funnel leaks at the last step. That is the whole reason my sister studio ClickTheory exists, because I got tired of watching good clip work die on a bad thumbnail.
How I would run it, concretely
If I were starting a video podcast tomorrow with limited hours, here is the plan I would follow.
- Record with clips in mind. Restate questions. Land points cleanly. Leave air after the good line.
- Pull four to six moments per episode, not twenty.
- Cut every clip for its platform, not once for all three.
- Shorts get a handoff line. TikTok clips do not need one. Reels get the guest tagged.
- Give it 90 days before judging anything. The first month tells you nothing.
- Track one number: full episode views or listens, not clip views.
That last one is the whole argument in a sentence. Clip views are a vanity metric right up until they turn into episode views, and the platforms differ enormously in how often that conversion happens.
None of this is expensive to do properly, it is just tedious, which is exactly the kind of work worth handing off. If you want to see what that looks like in practice, the full breakdown of how we run an episode covers where clips fit into the turnaround.
The short version
TikTok for discovery. Shorts for conversion. Reels for guests and for maintenance. Cut them differently, post fewer of them, and judge them on what happens to your episode numbers rather than what happens to the clip.
And if after three months the clips are working and the episodes still are not growing, the clips were never the problem. Go back and listen to minute 20 of your last episode with a stranger's ears on. That is usually where the answer is hiding.
If you want a second opinion on which moments in your show are actually clippable, send me an episode. I will tell you honestly if there are six in there or two.