Filler Words

Filler words are the little sounds and phrases that pad out speech without adding any meaning. The most common ones in English:

• Sounds. “Um,” “uh,” “er,” “ah.”
• Habit phrases. “Like,” “you know,” “I mean,” “right?”
• Repeated transitions. “So,” “and so,” “and then,” “basically.”
• Trailing phrases. “…and stuff,” “…or whatever,” “…you know what I mean?”
• Repeats. Saying the same word twice (“I-I think…”) while your brain catches up.

Why we use them: speech isn’t planned out in advance. While you’re talking, you’re also figuring out what to say next. Filler words buy you the half-second you need to line up the next sentence without going silent. They’re not laziness. They’re just a normal part of how people talk.

Filler words aren’t the same as rambling or fuzzy thinking. A clear thinker can still use plenty of fillers. A rambler usually has different problems: repeating themselves, going off on tangents, never quite landing the point.

In everyday conversation, filler words are invisible. In recorded speech (podcasts, video, voice-over), they suddenly stand out. The same sentence that sounded fine in person sounds halting and unsure on playback. That’s why cutting filler words is a normal part of video and podcast editing.

A few reasons filler words hit differently on video than in a live chat:

• No body language. In person, you read faces and gestures, and your brain fills the pauses with context. On video, especially with the camera tight on a face, every “um” registers.
• No back-and-forth. A real conversation has natural pauses and turns. In a one-way video (talking-head, voice-over), the viewer has nothing to do during a filler pause except notice it.
• Higher expectations. People expect recorded video to feel more polished than chat. The bar in their head is broadcast or scripted content, where the fillers are already gone.
• Repetition stands out more. In a 10-minute conversation, ten “ums” feel natural. In a 2-minute video, ten “ums” feel like a lot. How packed-together the fillers are matters as much as the total count.
• It costs you viewers. On YouTube and short-form, people click away when something feels slow or unsure. Filler words add to that. Cleaner audio holds attention longer.

The fix is editing. Modern software makes it possible to cut filler words in seconds, sometimes automatically.

There are two main ways to remove filler words in modern editing:

Cutting by hand. The editor scrubs through the audio, finds each “um” or “uh,” and cuts it out. Then they close the gap, joining the audio on either side, often with a tiny crossfade to smooth the join. Slow but precise. It’s been the standard for podcast and video editing for decades.

AI-assisted cutting. Tools like Descript, Premiere Pro’s Enhance Speech, and others can spot filler words on their own and remove them in one click. The AI writes out the audio as text, finds the fillers in the transcript, and trims the matching audio. Much faster, usually 80 to 90% accurate, and still better with a human checking the result.

The real trick to good filler-word cutting is keeping it sounding natural. A few rules of thumb:

• Don’t cut every single “um.” Some pauses are natural. Strip out every breath and hesitation and the speech starts to sound robotic.
• Cut the worst ones first. Long pauses, stacked fillers (“um, like, you know…”), and stutters are the priority.
• Smooth the joins. Two clean cuts back-to-back can sound abrupt. A short crossfade or a tiny silence between them sounds more natural.
• Watch the breaths. Cutting too tight can chop the start of a word. Leave a few frames of breath before each phrase.
• Keep the energy. Don’t cut so much that the speech feels rushed. A few natural pauses keep it sounding human.

For video, filler-word cutting usually comes with a bit of re-framing. Cuts in the audio show up as visible cuts in the picture, which can look jarring. So the editor adds B-roll, a slight zoom, or a graphic to hide the cut.

Editing fixes filler words after the fact, but cutting down on them while you record saves edit time. A few tips:

• Plan your key points. Walking in with a clear list of 3 to 5 points to hit means less thinking on the spot. Less thinking means fewer fillers.
• Pause instead of filling. Train yourself to take a breath instead of saying “um.” A silent pause edits out cleanly. An “um” takes more careful trimming.
• Record in shorter takes. Speaking for 30 to 60 seconds at a time, then resetting, is easier than going for five minutes straight. Fewer fillers per take.
• Listen back to yourself. Reviewing your own recordings makes you aware of your filler habits. Most people are surprised by how often they say them.
• Slow down. A lot of fillers come from talking faster than you’re thinking. Easing off the pace even a little cuts them down noticeably.
• Practice the opener. The first sentence is often the worst for fillers. Rehearse your opening line and the rest tends to flow more cleanly.

For interview guests who aren’t used to being on camera, expect more fillers. That’s normal. The fix is editing afterward, not asking them to perform like a TV host.