On this page
Write the product action and its visible result first. Only then choose the reaction and opening line that point toward them.
Hover or tap to see the motion. This is source footage, ready to combine with your own demo. Paid members can browse and download the originals.
What comes with a clipThe five-beat script you can fill in
Use this as an editing brief, not a formula for a guaranteed result. The timings are a first draft for a short vertical video. If the action needs more time to remain legible, extend it and remove another beat instead of speeding through the proof.
Fill the blanks with something your app actually shows: When [specific starting state], I can [one visible action] and end with [observable result]. Your opening line should create a question that the action answers. Your closing instruction should take viewers to that same task.
| Beat | Suggested timing | What to write or record |
|---|---|---|
| Reaction and question | 0–2 seconds | A human expression plus one short editorial line about the task. Do not present the person as a customer. |
| Starting state | 2–4 seconds | Show the real screen, item or problem. Make the relevant part large enough to read. |
| Product action | 4–9 seconds | Record the actual taps or clicks that perform one task. Keep any necessary step visible. |
| Observable result | 9–12 seconds | Hold on the changed state so a viewer can inspect it. |
| Next step | 12–14 seconds | Give one instruction that matches the demonstrated feature and destination. |
Example one: find a saved item
Imagine a saved-items app with a real search function. Populate a demo account with sample entries before recording. The point is not to show that search exists; it is to let a viewer follow a particular lost item all the way to the right result.
Use a curious or mildly surprised reaction, then caption it with “Where did I save that?” Cut to the saved list. Type the exact phrase you remember, show the matching result and open it. End on the item itself with “Try searching your saved items.” If the app only searches titles, do not imply it understands any phrase from the item's contents.
| Time | On screen | Caption or audio |
|---|---|---|
| 0–2s | Human reaction | Where did I save that? |
| 2–4s | Populated saved list | No extra claim; let the question carry. |
| 4–9s | Real query typed; matching result selected | Search the words you remember. |
| 9–12s | Correct item open and readable | There it is. |
| 12–14s | App or feature destination | Try it with your saved items. |
Example two: clean a messy CSV
For a data-cleaning SaaS, the useful proof is a before-and-after state that a viewer can inspect. Prepare a sample CSV with visible duplicate rows. Record the import, the real cleanup action and the resulting row count or reviewed output. Do not cut out a required confirmation step simply to make the tool seem automatic.
Use a puzzled reaction and the line “Why is this list twice as long?” Show the duplicate rows. Run the actual deduplication workflow. Pause on the reviewed output, then close with “See how the cleanup works.” Avoid a time-saved claim unless you have a measured, relevant comparison.
| Time | On screen | Caption or audio |
|---|---|---|
| 0–2s | Puzzled human reaction | Why is this list twice as long? |
| 2–4s | Sample CSV showing duplicate rows | The same entries appear again. |
| 4–9s | Actual import and cleanup action | Find duplicates, then review the change. |
| 9–12s | Cleaned list and visible review state | Now the list is easier to check. |
| 12–14s | Relevant product page | See the cleanup workflow. |
Example three: reschedule a workout
For a fitness app, film a real schedule change using sample data. The result must be visible in the calendar; a generic happy face or a tap on a button does not prove the workout moved. Choose a modest expression that suits a small, practical improvement.
Open with “Plans changed?” over a reaction. Show the workout in its original slot, move it using the app's actual control, and hold on the new slot. Close with “Move your next workout” only if the destination lets people do that. This is a proposed edit, not a user's story or an advertising result.
| Time | On screen | Caption or audio |
|---|---|---|
| 0–2s | Human reaction | Plans changed? |
| 2–4s | Workout in its original calendar slot | Here's today's plan. |
| 4–9s | Actual reschedule control in use | Move it to a time that works. |
| 9–12s | Workout visible in the new slot | The plan is updated. |
| 12–14s | Relevant app destination | Move your next workout. |
Give the editor a brief, not just a caption
Send the editor the starting screen, the exact action, the finished state, the chosen reaction filename and the closing destination. Note anything that must remain readable. Keep a continuous reference recording even if the final cut uses close-ups; it helps everyone check that a required step was not hidden.
Before exporting, play the sequence without sound at phone size. Ask whether the question receives an answer, whether the viewer can see the result and whether the face has accidentally become a testimonial. If the answer depends on narration alone, make the screen action clearer.
- Can the demonstrated feature actually perform the action in the script?
- Does the result stay on screen long enough to inspect?
- Does a caption imply that the filmed person used or endorsed the product?
- Are any time, money or outcome claims supported by evidence?
- Does the closing instruction lead to the feature just shown?
Questions & answers
Do I need a creator speaking the script?
No. These examples use a human-shot reaction as source footage, followed by your own product recording and editorial captions. If the video needs a person to describe their personal experience with your product, arrange a suitable product-specific shoot.
Can I use the example wording exactly?
Yes, where your product really performs the action shown. Treat the wording and timings as an editable brief, not as tested advertising claims.
Find the opening for your next edit.
See the public samples, then unlock the library when you’re ready to work with the originals.
Explore the library