LearnDeliver it
Do your programs need exercise videos?
The short answer
Yes, for any exercise the person might not already know. In a study of a five-exercise program, people given written instructions alone made more than twice as many errors as people shown the movements live or on video, and video did as well as a live teacher. A short, plain clip is enough. You do not have to film every exercise yourself.
2x
the errors when people learn from written instructions alone1
4.5x
the effect of a demo on form, compared with its effect on results2
No gain
from slow motion over real speed, in a trial of 51 people3
A program is a set of instructions for moving. Words are a poor way to describe movement, and the person reading them is usually alone, mid-session, with nobody to ask. A demo closes that gap. The useful questions are how much it helps, where it does not, and how little effort a good one takes.
What does a demo actually change?
It changes how well people do the movement.
- Written instructions alone produce more errors. Forty adults learned five shoulder exercises from a live instructor, a video or a written handout. The handout group made more than twice as many errors as the others. Two video versions were tested, one showing only correct form and one showing mistakes being corrected. Both did as well as the live instructor.1
- Demonstration improves form more than results. A meta-analysis of observational learning found that watching a model had a large effect on how a movement was performed (0.77) and a small one on the outcome it produced (0.17).2 That fits exercise well. Good form is the goal of a demo.
- It helps with lifting, on top of coaching. Fifteen novices learning the power clean improved their technique faster when they watched a video of a skilled lifter before each set, in addition to normal coaching cues.4
- The finding has held up. A review of 48 studies from 2011 to 2018 confirmed that watching demonstrations works for applied motor skills, across a wider range of people and settings than earlier work had covered.5
- How the movement is performed0.77
- The result the movement produces0.17
Meta-analysis of observational learning, against practice alone. By convention an effect of 0.8 is large and 0.2 is small.2
Do videos make people stick to a program?
There is no good evidence that they do. This is the claim to be careful with.
| Study | Who | What it found |
|---|---|---|
| Meta-analysis, 2019 | Rehabilitation patients | Multimedia instructions "may" improve adherence over written ones, on very low-quality evidence. No gain in patient outcomes6 |
| Randomised trial, 2017 | 62 people after a stroke | Filmed home exercises with reminders did no better than paper for adherence7 |
| Randomised trial, 2017 | 80 people with joint or muscle problems | A digital program with phone calls and text messages beat paper handouts by 1.3 points on an 11-point adherence scale8 |
The third study is the positive one, and it bundled video with human follow-up, so the video cannot take the credit. The honest summary: a demo helps people do an exercise properly. Whether they turn up is decided by other things, which are covered in why clients stop following their program.
Which exercises need a video?
Not all of them. Spend the effort where a name alone would send someone to a search engine.
| Needs a demo | Usually fine without |
|---|---|
| Anything with a name you made up or shortened | Walking, cycling, rowing machine |
| Hinges: Romanian deadlift, kettlebell swing, good morning | Fixed-path machines with a diagram on them |
| Unilateral work: split squats, single-leg deadlifts | Planks and holds the person has done before |
| Olympic lift variations and anything explosive | A second or third appearance of an exercise already shown |
| Mobility drills and warm-up flows | |
| Your own variation of a common exercise |
A simple test: hand the exercise name to someone who trains a little. If they hesitate, it needs a demo.
Do you have to film them yourself?
No. There are three ways to put a demo next to an exercise, and each has a cost.
| Option | Good for | The cost |
|---|---|---|
| Film your own | Your variations, your cues, your face. It is the part of a program that feels like you | Time. A full program can mean 40 to 60 clips |
| Use a demo library | Standard exercises, shown consistently, with no filming | It is not you, and a library may not have your variation |
| Link to a public video | Nothing to make | The person leaves the session, may meet adverts, and the link can break or change |
Most coaches end up mixing the first two: a library for the standard lifts, and their own clips for the ten or so movements that make the program theirs.
How do you film a demo with one phone?
The research is reassuring here. Most of what people worry about does not matter much.
- Angle is not worth agonising over. In one small study from 1995, a model filmed from behind was learned faster at first than a mirrored front view, but people remembered both equally well.9 Pick the angle that shows the joint that matters, and keep it.
- You probably do not need slow motion. A trial of 51 people learning a dance skill found that slow-motion, real-time and combined demonstrations produced the same learning.3
- Showing the mistake can help. In a lab study, watching both an expert and a novice produced better learning than watching either alone.10 For a movement people often get wrong, one clip of the common mistake beside the clean version is reasonable.
- Short and informal is fine. A study of 6.9 million sessions of instructional video found people engaged more with short videos, and that informal filming did as well as studio production.11 That was university teaching, not exercise, so take it as a direction.
Filming checklist for one phone
How DUAL handles demos
Most exercises in a DUAL workout are matched to a demo from its exercise library, so a program is followable the day you publish it. You can film your own clip for an exercise and show that instead. Your coaching stays on screen in every session, long after the post has scrolled away.
See a converted workoutQuestions people ask
Long enough to show two or three clean reps, which is usually under 15 seconds. There is no research on the ideal length of an exercise demo. Studies of instructional video point the same way: shorter clips are watched more fully.
No. A silent clip that shows the movement clearly does the main job, and many people train with their own audio on. One short cue, spoken or as a caption, is a useful extra.
They do not need to be. The one trial that compared slow-motion, real-time and combined demonstrations, for a dance skill, found no difference in learning. Real speed also shows the tempo you want.
Only with permission or a licence. A video being public does not make it free to reuse in something you sell. Film your own, or use a library that licenses its demos for this purpose.
Sources
Every source was read and checked on 11 November 2026. Spotted something out of date? Tell us.
- Reo JA, Mercer VS. Effects of live, videotaped, or written instruction on learning an upper-extremity exercise program. Physical Therapy, 2004.
- Ashford D, Bennett SJ, Davids K. Observational modeling effects for movement dynamics and movement outcome measures across differing task constraints: a meta-analysis. Journal of Motor Behavior, 2006.
- Lelievre N, St Germain L, Ste-Marie DM. Varied speeds of video demonstration do not influence the learning of a dance skill. Human Movement Science, 2021.
- Sakadjian A, Panchuk D, Pearce AJ. Kinematic and kinetic improvements associated with action observation facilitated learning of the power clean in Australian footballers. Journal of Strength and Conditioning Research, 2014.
- Ste-Marie DM, Lelievre N, St Germain L. Revisiting the applied model for the use of observation: a review of articles spanning 2011-2018. Research Quarterly for Exercise and Sport, 2020.
- Emmerson KB, Harding KE, Taylor NF. Providing exercise instructions using multimedia may improve adherence but not patient outcomes: a systematic review and meta-analysis. Clinical Rehabilitation, 2019.
- Emmerson KB, Harding KE, Taylor NF. Home exercise programmes supported by video and automated reminders compared with standard paper-based home exercise programmes in patients with stroke: a randomized controlled trial. Clinical Rehabilitation, 2017.
- Lambert TE, et al. An app with remote support achieves better adherence to home exercise programs than paper handouts in people with musculoskeletal conditions: a randomised trial. Journal of Physiotherapy, 2017.
- Ishikura T, Inomata K. Effects of angle of model-demonstration on learning of motor skill. Perceptual and Motor Skills, 1995.
- Rohbanfard H, Proteau L. Learning through observation: a combination of expert and novice models favors learning. Experimental Brain Research, 2011.
- Guo PJ, Kim J, Rubin R. How video production affects student engagement: an empirical study of MOOC videos. ACM Learning at Scale, 2014.