How to weave two testimonials into one story
Upload both interviews and tell Vizard Agent you want one story rather than two testimonials in sequence. Vizard Agent transcribes all of it, finds where the two accounts agree and where they add something different, pulls the official branding, and cuts them into a single argument ending on the ask.
What is the short version?
Two people saying the same thing one after the other is twice as long and no more persuasive. Cutting between them so each picks up where the other left off makes one case rather than two, and it is shorter.
- Go to Vizard Agent and upload both interviews.
- Say you want one woven story, not two testimonials in sequence.
- Say what the video is asking people to do.
What do you need before you start?
Both interviews in full and a clear ask. Vizard Agent reads everything and finds where the two accounts intersect, so no logging is needed from you. Knowing what the video is actually for is what decides which parts of each interview matter.
- Both interviews. Unedited, in full.
- The ask. Apply, attend, donate, book.
- The organisation. So its branding can be pulled.
- The length. A minute or two.
- Consent. For this specific use.
What do you type into Vizard Agent?
Say "one narrative" explicitly. Vizard Agent otherwise delivers the safe reading — the first person's testimonial followed by the second's — which is what most edits of two interviews look like and is exactly the thing worth avoiding.
Prompt
Variants worth knowing:
- Interleaved rather than sequential. Each voice advancing the other.
- Official branding pulled. From the organisation's own material.
- A single ask. At the end, once.
What does Vizard Agent actually do?
Here is the order Vizard Agent worked in on a real two-testimonial edit for an event organiser. The reading comes first throughout: all three source videos transcribed in full before any decision at all about which lines to use.
- Loads the project ledger, the skills and the asset catalogue.
- Loads the customer-story editing guidance.
- Probes all three source videos and starts extracting frames.
- Checks the transcription tool and extracts frames from all three interviews.
- Transcribes all three interview clips in full.
- Downloads the organisation's existing brand assets.
- Looks up the official vendor branding and the call to action for that organisation.
- Extracts real frames at multiple timestamps and resizes the brand images for review.
- Builds contact sheets of both interviews, then reviews the previous logo asset, the brand photographs and the earlier advertisement's style.
Step 7 matters more than it looks. The exact wording of a call to action is usually published by the organisation, and using their phrasing rather than an approximation is what makes the ending sound official rather than improvised.
What does the result look like?
The interviews this was built from measure 1280x720 at 30fps, one running 107 seconds. From two of them comes a single story: one speaker opening the point, the other confirming it from a different angle, the organisation's branding throughout, and one ask at the end.
Neither testimonial is complete on its own in the finished cut. That is what "woven" means — each speaker appears where they add something, rather than each getting their turn.
When does this not work well?
Interleaving two accounts assumes they genuinely belong in the same argument, and sometimes they simply do not. These are the cases where Vizard Agent will say so plainly rather than smoothing two genuinely different experiences into a single one.
- Contradictory accounts cannot be woven. Two different experiences are two videos.
- Editing must not distort. Cutting between speakers must not imply agreement that is not there.
- Consent covers a use. An interview given for one purpose does not cover another.
- Two voices need distinguishing. Similar speakers confuse a viewer when interleaved.
- One strong testimonial beats two weak ones. More material is not more persuasive.
How do you fix a result that came back wrong?
Say which passage or which speaker came out wrong, and what you wanted there instead. Vizard Agent keeps all three transcripts, the contact sheets, the brand assets and the looked-up wording, so the weave changes without re-reading anything.
- "The second speaker gets lost." Re-balanced across the story.
- "That implies she said something she did not." Re-cut to keep the accounts separate.
- "Use their official wording." Rebuilt from the looked-up call to action.
How does Vizard Agent compare to doing it yourself?
By hand this means transcribing two whole interviews, hunting through both for the lines that answer each other, and then building a timeline that alternates between them — all while making sure the interleaving never puts words into anybody's mouth.
| By hand | Vizard Agent | |
|---|---|---|
| Reading | Watch both, twice | Both transcribed in full first |
| Structure | One then the other | Interleaved into one argument |
| Branding | Whatever you have saved | Pulled from the organisation's own material |
| The ask | Written from memory | Looked up in their published wording |
Common questions
Why weave rather than play them in turn? Because two people saying the same thing consecutively is twice the length and no more convincing.
How many testimonials can it combine? Two works cleanly for Vizard Agent. Three starts to need distinguishing devices for the viewer.
Will it misrepresent anybody? It should not, and you should check. Interleaving must not manufacture agreement.
Can it use official branding? Yes. Vizard Agent pulls the organisation's own assets and its published wording.
How long should it run? A minute or two. Long enough for both voices, short enough to end on the ask.
What if the two accounts disagree? Then they are two videos. Vizard Agent will say so rather than smoothing the difference away.
Do I need consent from both? Yes, and for this use specifically. A recorded interview is not blanket permission.
Should the ask appear more than once? No. Once, at the end, after both voices have made the case.
Can it caption them? Yes. Vizard Agent builds them from the transcripts it already has.
Can a third voice be added later? Yes, though it changes the shape. Vizard Agent can weave a new interview into the existing structure, but three voices need clearer visual separation than two do.
Should both speakers appear on screen together? Not usually. Vizard Agent cuts between them, because a split screen makes the two accounts look like a debate rather than one argument built from two people.
How long should each speaker be on screen? In turns of a few seconds. Vizard Agent alternates often enough that the two feel like one voice arguing, and rarely enough that nobody gets cut off mid-thought.
Does this suit fundraising as well as recruitment? Yes, and better in some ways. Vizard Agent builds the same shape — two accounts, one case, one ask — and a cause argued by two beneficiaries reads as evidence rather than as a claim.
What if one interview is much better than the other? Say so, and lead on it. Vizard Agent still uses the weaker one where it adds something the stronger account does not cover, rather than giving both equal time on principle.
Does Vizard Agent check the result? Yes. It reviews the cut and the branding before delivery.