When to consider removing ums, ahs and filler words from a Podcast

There is a perception that cleaning out filler words and hesitations from a conversation is an easy win, making an amateur host or guest sound more "professional". I think that perception is entirely wrong.

It's a request I come across occasionally with podcast clients. "Can you remove uhms and ahs from my podcast."

I would define filler words as anything that is not essential to understanding a conversation but is also entirely natural to it. 

Um, ah, eh, like, right, y'know

Have you ever considered that your listener doesn't want polished and professional?

Podcasts are a window into real-life conversations. That is what makes the medium so good and separates it from the polished nature of old-fashioned mediums like TV and radio.

Over-editing can not only kill this beautiful rawness, but it can also sound terrible if not done correctly.

With everything you do in a podcast, you need to ask why. Why are you removing filler words, or why do you want them removed?

In rare occasions, removing filler words is a good approach, but consider the following before deciding it's the best route for your podcast.

Focus on more important things first.

What gets me is when people ask for filler words removed when they get so many more essential things wrong. Recording terrible audio and requesting to remove filler words is a complete no-no in my book.

Filler words are the final polish you should only consider when you conquer some basics first:

• Record with good equipment.

• Learn how to use the equipment correctly and guide your guests with those learnings.

• Nail down a structure and format that is slick.

• Use a remote recording platform like Riverside.fm

• Remove other larger chunks of content first

If you consider removing filler words before any of the above, you are trying to run before you can crawl.

In a nutshell, please don't consider removing ums and ahs when you record your podcast with an internal laptop microphone on zoom. I see this a lot, and it's entirely counterproductive. 

Similarly, filler words should not be the first stop in the editing process. If you want to edit your podcast down, edit the conversation first. Take out unnecessary chunks or boring parts, then move on to minor elements.

It's time-consuming and costly.

Editing yourself will take a long time, especially if you aren't an experienced audio editor.

I charge double the rate to remove filler words and only do it when clients master the recording and format. It is really time-consuming for an editor, and the payoff is (in my opinion) zero to minimal.

Do you really think editing uhms and ahs makes your podcast better, and why?

Is it worth the extra time or cost? Nearly always no.

It won't make the podcast noticeably shorter.

Having edited tons of podcasts, I can tell you that removing filler words will save an average of 2-3 minutes on a 40-minute podcast. 

I greatly advocate shorter podcasts, but a podcaster would be better to remove large chunks of less interesting content than focus on filler words.

It won't impress the listener or make a podcast sound more professional.

If you want to sound slicker and become a better podcast host or guest, then you should invest in some coaching with a broadcast coach (I know a few who are excellent).

Filler words and gaps give context to a conversation or story.

If a host asks a killer question and the guest pauses, then says "ehhh", it gives context, shows the thought process and lets the listener know that this isn't a simple, straightforward answer.

This scenario is good, and you shouldn't want it removed from a podcast. Some of the best moments in any interview come from moments of silence and hesitations.

AI “remove filler words” tools sound awful.

AI is fantastic for certain things, but I'm yet to hear the evidence on this one. Descript and a couple of other tools offer this, and I think it should come with a warning.

I think they are okay on short clips for social, but on a whole 40-minute conversation, they make a podcast sound worse.

In seconds of listening to a podcast, I can tell when AI has been used to remove filler words. You can always hear half-breaths and blips that are entirely unnatural.

Anyone who edits audio will tell you there is a specific science to editing words and making it seem seamless. Sometimes it's impossible because filler words are joined with other words. When an AI tool tries to tackle these moments, it sounds horrendous.

AI is improving rapidly, but it cannot remove filler words while maintaining the natural human sound.

Times when you should consider removing filler words.

On the whole, I try and push people away from the notion of removing filler words. Most podcasts are casual chats and should be left that way.

However, here are the times I think it's worth polishing up those filler words:

Intros and Outros

As intros and outros are not part of the actual conversation, there is more of a case for these moments to be polished up and made to sound slick.

The first couple of minutes of a Podcast

Not essential, but if you want to sound slicker, a clever trick is to remove filler words in the first couple of minutes. 

These are the key moments when the listener will decide whether or not they want to stay with the podcast. They are also moments when a host or guest can be nervous and not yet in flow.

So, if the early stages of a podcast sound patchy with uhms and ahs, then it might be worth cleaning up those awkward first 2 or 3 minutes.

Narrative Podcasts

Documentaries or anything where the host speaks a narrative should have uhms and ahs cleared up. In this instance, it's an entirely different skill, so if you're a narrative-style podcast host, you should also double down on practicing narrative reads.

The host or guest is insecure about a verbal tick or stuttering.

I have seen this instance where a host or a guest specifically asks for verbal ticks or stutters to be removed. This request might be due to an insecurity with a tick or verbal disorder. 

The final polish

When everything else is right (audio, story, format, structure) and you have first done an editorial, choose to remove the conversation's actual chunks, making the overall podcast shorter and more impactful. Then as a final pass removing filler words is not a bad idea.

Here is an example of a podcast in which I do this. This podcaster worked long and hard with me on nailing down audio quality, structure and hosting skills before we got to this point: