EchoPitch Statistics
Original research data from EchoPitch AI analysis. These statistics are derived from presentation recordings and can be cited under CC BY 4.0.
Quick Answer
EchoPitch's statistics page presents original research data on presentation anxiety, gathered from AI analysis of practice recordings, alongside citable external peer-reviewed research. Speakers self-report anxiety at 8-9 out of 10, but AI facial analysis using the Facial Action Coding System measures visible anxiety at only 3-5 out of 10, a 40-60% perception gap between how nervous you feel and how nervous you actually appear to an audience. The data also covers vocal delivery: professional presenters average 130-150 words per minute, while anxious speakers average 170-200 WPM without noticing they're rushing. Additional statistics cover practice effects, including the 3-Rep Rule for deliberate practice and the first-30-seconds credibility judgment window. Original EchoPitch statistics are licensed under CC BY 4.0 and freely citable; external statistics, such as the 75% prevalence of public speaking anxiety, credit their original peer-reviewed academic sources instead.
How to cite
EchoPitch. (2026). [Statistic name]. Retrieved from https://echopitch.io/statistics
Licensed under Creative Commons Attribution 4.0
Perception Gap
Average self-reported anxiety
Speakers consistently rate their internal anxiety at 8-9 out of 10 before/during presentations.
Methodology: Self-report questionnaire administered immediately before recorded practice sessions.
AI-measured visible anxiety
AI facial analysis measures visible anxiety signals at only 3-5 out of 10 for the same speakers.
Methodology: Facial Action Coding System (FACS) analysis via computer vision, measuring Action Units associated with anxiety (AU1, AU4, AU15, AU20).
Perception gap
The difference between internal experience and external appearance is consistently 40-60% — your audience sees far less anxiety than you feel.
Methodology: Calculated as the difference between self-reported anxiety and AI-measured visible signals, expressed as a percentage.
See full context →Vocal Delivery Patterns
Optimal speaking pace
Professional presenters maintain 130-150 words per minute. Below 120 signals disengagement; above 160 signals nervousness.
Methodology: Analysis of speech transcription timestamps across recorded presentations.
See full context →Anxious speaker average
Most anxious presenters average 170-200 WPM — fast enough to signal anxiety but not so fast they notice they're rushing.
Methodology: Measured pace among speakers self-reporting anxiety scores of 7+ out of 10.
See full context →High hesitation density threshold
Filler words, false starts, and mid-sentence pauses exceeding 5 per minute signal uncertainty to audiences. Professional speakers average 2-3.
Methodology: AI transcription with filler word detection and categorization.
See full context →Practice & Improvement
The 3-Rep Rule threshold
Performing a skill correctly 3 times consecutively indicates learning rather than luck. This applies to openings, transitions, and closes.
Methodology: Deliberate practice methodology based on motor learning research (Ericsson).
See full context →First impression window
Audiences form credibility judgments in the first 30 seconds, before consciously evaluating content. Opening delivery determines how the rest is perceived.
Methodology: Based on primacy effect research and vocal credibility signal literature.
See full context →Mental practice activation
Mental rehearsal activates motor cortex at 60-70% of physical practice levels, making visualization a valid complement to recording practice.
Methodology: Driskell et al. (1994) meta-analysis on mental practice and motor performance.
See full context →External Research Statistics
These statistics come from peer-reviewed research. Please cite the original sources, not EchoPitch.
of people experience some degree of anxiety about public speaking
National Comorbidity Survey Replication (Kessler et al., 2005)
DOI: 10.1001/archpsyc.62.6.593most common social phobia — ranks above fear of death in prevalence
Stein & Stein (2008), The Lancet
DOI: 10.1016/S0140-6736(08)60488-2respond well to systematic exposure-based practice
Hofmann & Smits (2008), Dialogues in Clinical Neuroscience
DOI: 10.31887/DCNS.2008.10.2/sghofmannFull citations available at /research