The parts of a talk you can actually count
Feedback after a presentation arrives as one word. "Confident." "A bit flat." A listener experiences delivery as a single impression, so the note you get back is unusable as practice material. Underneath that impression sit separate behaviors that leave a trace on the recording: your rate in words per minute, which kind of filler you reach for, how many words you spend before the first claim lands, how often you hedge a sentence you meant to assert.
The Wellspoken Index splits a recording into six of these and weights them: Structure 250, Conciseness 200, Confidence 150, Pronunciation 150, Filler Rate 150, Pace 100. The split matters more than the weights. These dimensions move independently, so one speaker holds a textbook pace while hedging every claim, and another is organized and crisp and still burns forty seconds before arriving at the point.
Rate is the first number worth checking
| Range | How it lands |
|---|---|
| Under 120 wpm | Deliberate. Suits heavy or emotional material, drags on lighter topics. |
| 120 to 150 wpm | Measured. The range most coaching aims for. |
| 150 to 170 wpm | Brisk. Works when you pause at transitions. |
| 170 to 190 wpm | Fast. Followable, but only with real gaps between ideas. |
| Over 190 wpm | Very fast. Hard for note-takers and second-language listeners. |
Above 190, the fix is usually pausing rather than slowing down. The words themselves are fine, the gaps between ideas have been deleted, and the audience never gets a place to catch up. Slowing every individual word feels artificial and rarely survives contact with a real room. One full beat of silence at each section boundary does the same work and holds up better under pressure.
Time yourself with the pace test. For a fixed slot, the speech time calculator runs the arithmetic at your rate rather than a generic one: a script that fills twenty minutes at 130 wpm finishes in about fifteen at 175. On picking a target, see how fast you should talk.
Fillers come in four kinds with four different fixes
Counting "um" catches a fraction of the problem. The filler word counter separates:
- Hesitation sounds (um, uh, erm, hmm). Pure stalling while the next phrase assembles. The group that replaces most cleanly with silence.
- Discourse fillers (like, you know, I mean, sort of). Real words used as padding between ideas. Harmless in small numbers, distracting once they cluster.
- Qualifiers (basically, actually, literally, honestly, just). These usually add nothing to the sentence they sit in, and cutting them tightens delivery immediately.
- Sentence crutches (so, right, okay, well, anyway). Openers and closers running on autopilot. Most speakers have exactly one and cannot hear it.
Run a transcript through and the distribution is often not the one you expected. The "um" habit you came in worried about turns out to be a qualifier habit, which needs a rewriting drill rather than a pausing drill. Replacing hesitations with silence and cutting a ramble down get confused with each other constantly.
What changes once you have numbers
Practice gets specific. Rather than running the whole talk again and hoping, you work the dimension that scored lowest and re-record to see whether it moved. A structured answer format fixes a structure score and does nothing for a pace score, and the number tells you which problem you have.
Measurement also changes what you should expect of yourself over time. Career-span survey data covered in our review of communication apprehension at work found that overall apprehension and public-speaking apprehension both declined with age, years of full-time work, and how frequently people gave presentations. Frequency was one of the variables that moved. Reps are the treatment, and a score is what turns a rep into a targeted one.
The annotated speaker breakdowns work the same way. William H. McRaven structured a 2014 commencement address as ten numbered lessons, each closing on the same short imperative, and scores 235 out of a possible 250 on Structure in our reading of it. Simon Sinek and Amy Cuddy get the same line-by-line treatment. Reading a transcript with the metric named beside the sentence teaches more than watching the talk and admiring it.
