[FAQ]

I’m in the middle of editing my podcast episodes, and I’m going back and forth on what to do with all the “ums,” “ahs,” and other filler words from the speakers.

On one hand, I want the audio to sound as clean and professional as possible. On the other hand, I’m worried that if I remove every single one, the conversation will sound robotic and lose its natural, human feel.

What’s the general best practice for this in 2025? Is there a rule of thumb for how many to cut versus how many to keep to maintain a natural, conversational flow?

Curious to hear how other editors approach this. Thanks!

[Jeff Bullas]

This is a classic editing dilemma for anyone working with spoken-word audio. The best approach is not about total removal, but about finding the right balance between sounding professional and sounding authentically human.

The goal is to improve the clarity and flow of the conversation for the listener, not to achieve robotic perfection. Here are the best practices for handling filler words:

Remove Distracting Fillers: You should focus on editing out the filler words that are most noticeable and distracting. This includes long, pronounced “umms” and “ahhs,” especially those that occur during a long pause while the speaker is thinking. You should also remove false starts or repeated words that make a sentence confusing to follow.

Keep the Natural-Sounding Fillers: It is often best to leave in the very short, almost unnoticeable filler words that occur within the natural flow of a sentence. Removing every single one can make the speech sound jarring, choppy, and overly edited. A conversation without any imperfections can lose its human, relatable quality.

Prioritise Clarity Above All: The guiding question should always be: “Does this filler word hurt the clarity of the sentence?” If the answer is yes, remove it. If the answer is no, and removing it makes the speech sound unnatural, it is often better to leave it in.

Be Mindful of AI Tools: Many modern editing tools, such as Descript or Adobe Podcast, have excellent AI features that can automatically detect and remove filler words with a single click. While these tools are incredibly powerful and time-saving, it is still wise to review their automated edits. Sometimes they can be too aggressive and remove words that are needed to maintain a natural conversational cadence.

In summary, the goal is to make the speaker sound like a clear and confident version of themselves, not a robot. Remove the filler words that are distracting, but keep the subtle ones that preserve a natural, human flow.

Cheers,

Jeff

[Author]

Posts