Reducing them is great, but elimination isn’t necessary and may even backfire. 

If there’s one thing people know about public speaking, it’s that Fillers Are Bad: um's & uhs, like’s & kinna’s, right’s & so’s—they’re verbal clutter that make you look dumb & should be eliminated one and all. New students who have internalized this conventional wisdom are therefore surprised at my comparatively soft-on-fillers approach. Now, don’t get me wrong, I monitor them & regularly help students cut back. But in my view, a smattering usually doesn’t matter—and in some settings, they may actually improve how you sound. 

Because there’s nothing like an official-sounding Framework to make an idea sound impressive, allow me to introduce ...

1. The Threshold Theory of Speaking Imperfections

 Aside from content, audiences primarily notice repetition—and are eventually bothered by it. If, say, a speaker adjusts his glasses once or twice in a 5-minute speech, no one will be the wiser. If he does it every 15 seconds, the audience will soon be able to think of nothing else. So there’s clearly a continuum here: As a speaker does or says something more & more frequently, first it becomes noticeable to the audience, then distracting, & then finally annoying. But! There is a number that is above zero but below noticeable, where the number might as well be zero. All a speaker needs to do is stay below that threshold, and they’re golden. The audience will neither notice or care about those imperfections, & will instead stay focused on the ideas. 

 Getting those ideas across to the audience is usually a speaker’s top priority, and rhetorical flaws basically all fit into two (somewhat overlapping) categories: First, distracting from the ideas (e.g., by twirling a pen constantly); & second, making it harder for the audience to understand those ideas (say, by mumbling indistinctly). Fillers, unless egregious, don’t impair audience comprehension—they can still catch all the words & follow everything you say. So it doesn’t make sense to stress out over a couple of ums, in my view. 

In contrast to fillers, repetition falls under the Distraction category—and its effect can be strong enough to transform even good aspects of speaking into big negatives. For example, someone may have a go-to hand gesture that looks great! But as they return to it over & over, it draws more & more attention, until the audience is sick of it. The lesson: Variation, not perfection, is central to effective presenting.

So that’s how I recommend thinking of fillers: Not so much as Intrinsically Bad, & more as a potential issue that you don’t want to let get out of control. Plus, the (probably substantial) time, effort & angst usually required to stamp out Every Last Trace of fillers or other flaws usually isn’t worth it. There’s no discernible improvement in audience experience, & it likely comes at the expense of working on other, more impactful facets of one’s speaking.  

2. When fillers can actually help

Audiences generally appreciate authenticity—the sense that the speaker is being present with them, using their own words, and thinking up those words more or less in the moment.* Sure, they may be following an outline of some sort, and they don’t have to be presenting wholly original, personally meaningful ideas. This is fine. But even if you’re giving a simple work presentation, reading straight from the slides or a script probably won’t be very well-received. Lacking the organic cadences & intonations of ordinary speech, it will likely sound stiff, clunky and inauthentic. The phrasing will also probably be inferior, since sentence structures & word choices are typically different between speaking & writing. Put it all together, & the odds of sparking “this coulda been an email” grumblings are high. 

This is why, for a highly rehearsed or memorized speech,** adding a select few fillers or other minor imperfections can actually make you sound more natural—and as explained above, audiences won’t pick up on the flaws as long as they’re below the Noticeability Threshold. 

By way of example: I hosted a scripted podcast for a number of years, but wanted the episodes to sound like I and my guests were thinking up the content up live in-studio as we recorded. (I’m pretty sure many podcasts do it this way.) So I would write the occasional “um” or “like” or other mild verbal stumble into our scripted lines to give them that “we’re making it up as we go!” flair.*** I wanted the show to have character—and saying something has character is just another way of saying it has flaws. (🤯)

 A couple caveats: Not many folks I coach are giving speeches that run the risk of being over-rehearsed to the point of stiffness, and for those that do, there’s more to preserving a natural style than sprinkling in the odd “mistake” or two. That said, I think the fact that the much-maligned filler can actually be a boon in some situations helps dispel the black-and-white view many hold about it. 

*Heartfelt Scripted Speeches like wedding toasts get a pass because what they lack in in-the-moment articulation, they can make up for sincere feeling, humor, & vulnerability. Also, reading is pretty normal in those settings. 

**Less rehearsed speeches probably won’t be so uber-polished, so they wouldn’t have the inauthenticity problem. They may have other problems, but not this one. 

*** This is basically Lite Acting.