Vizard Agent

How to explain a job you finished to your customers

Last updated 2026-08-30 · 7 min read

Upload the footage from the job and tell Vizard Agent it is for customers rather than for the trade. Vizard Agent shot-logs the site footage, transcribes what was said on the day, finds the before and after photographs inside it, writes narration a customer can follow, and measures it against your target length.

What is the short version?

Trades film their work constantly and almost never publish it, because the raw footage is muddy, unnarrated and full of jargon. Turning it into something a customer understands is mostly a translation problem rather than an editing one.

  1. Go to Vizard Agent and upload the footage from the job.
  2. Say it is for customers, not for other tradespeople.
  3. Say the length — a minute is right for this.

What do you need before you start?

The footage from the day and the job's name. Vizard Agent reads what was said on site and finds the useful frames, so the recording does not need to be tidy, but the work has to be visible somewhere in it.

What do you type into Vizard Agent?

Name the job and the audience. Vizard Agent writes very differently for a customer than for a trade audience — no jargon, an explanation of why the work was necessary — and stating that is what stops the narration sounding like a method statement.

Prompt

Explain [the job] in a [60] second narrated video — for customers. Plain language, no jargon, show the before and after. [Attach the footage and any photos.]

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real job explainer for a drainage contractor. The step worth borrowing is the nineteenth: the before and after photographs were inside the footage, and it pulled them at full resolution rather than using the version that flashed past.

  1. Checks the workspace and the project history, then looks at the attached image.
  2. Probes the uploaded video and gets a shot log of the footage.
  3. Transcribes the spoken audio from the site.
  4. Pulls frames from the footage and reviews the frame grid.
  5. Checks the before and after photographs that appear inside the footage.
  6. Reads the available tool options and checks the key detail frames.
  7. Generates the voiceover and measures the narration's length.
  8. Searches for a music bed and generates the narration sections and the music together.
  9. Locates the generated audio, measures the duration and loudness of every stem, samples the opening and the photo section, then pulls the before and after photos at full resolution and reviews them.

Step 9's ending is a small thing that changes the finished video. A photograph shown on a phone screen inside a site video is unusable at full frame; the same photograph pulled out at its real resolution is the strongest shot in the piece.

What does the result look like?

The source footage this was built from measures 1280x720 at about 30fps and runs 120.15 seconds. From it comes a sixty-second narrated explainer: the problem, why it needed doing, the work itself, and the before and after shown at full quality with a music bed underneath the voice.

The narration explains rather than describes. That is the difference between a video that wins the next job and one that shows a trench to people who do not know what they are looking at.

When does this not work well?

The footage has to actually contain the work, and sixty seconds is not very long for a complicated job. These are the limits Vizard Agent runs into, and most of them are decided on site rather than in the edit.

How do you fix a result that came back wrong?

Say which part a customer would not follow, and why. Vizard Agent keeps the shot log, the whole transcript, the extracted photographs and the measured narration timings, so the script can be rewritten without re-reading any of the footage.

How does Vizard Agent compare to doing it yourself?

By hand this means watching your own site footage back, writing a script, recording narration in the van, and then finding that it runs ninety seconds. Most trades give up and post photographs instead, with no explanation attached.

By hand Vizard Agent
The footage Watch it back yourself Shot-logged and transcribed
The photos Screenshot from the video Pulled at full resolution
The script Write and re-record Written to a measured length
Language Whatever you say on site Rewritten for a customer

Common questions

Does my footage need to be tidy? No. Vizard Agent shot-logs rough site footage and finds whatever is usable inside it.

Can it use photos that are only inside the video? Yes. Vizard Agent pulls them out at full resolution rather than screenshotting the screen.

What about the noise on site? Vizard Agent replaces unusable site audio with narration and a music bed.

Will it use trade language? Not if you say it is for customers. That one instruction changes how Vizard Agent writes the script.

How long should it be? Around a minute. Long enough for Vizard Agent to explain one job properly.

Why explain rather than just show? Because a customer looking at a trench does not know what was wrong or why it mattered. Showing the work proves you did it; explaining it is what makes them call you.

Can I make one for every job? Yes, and that is where it pays off. Vizard Agent turns a library of finished jobs into a portfolio.

Should I appear on camera? Not necessarily. Narration over the work is usually easier and sounds more confident.

Can it do a shorter social cut? Yes. Vizard Agent cuts one from the same shot log and photographs.

Should the video name a price? Rarely. Vizard Agent will include one if you insist, but a job explainer works by demonstrating competence, and a number invites comparison before anyone has understood what the work involved.

What should the first five seconds show? The problem, not the company. Vizard Agent opens on what was wrong because a customer recognises a blocked drain instantly and has no reason to care about a logo until they do.

Does Vizard Agent check the finished video? Yes. It reviews the photo section and the audio levels before delivery.