A catalogue too big to film

The problem

Conventional production does not scale to a catalogue. Each video is a shoot, an edit and a round of revisions, so the cost per product barely falls no matter how many you make, and the ones that never get filmed are exactly the products nobody knows how to install. Getting it wrong is worse than saying nothing: an installation video is a technical document, and a claim on screen that is not in the literature is a claim the manufacturer has to stand behind. So the problem is not making video cheaply. It is making video that is both repeatable and accountable.

What we built

• Built a shared engine every video type imports, so adding a format is a new pipeline folder rather than a new production process
• Made the reel itself data: one config file carries the voice, the brand palette and fonts, the music ducking, the caption dictionary and the beats — twelve of them on the first one
• Synthesised the narration from the script, then force-aligned it against the generated audio so captions land word by word instead of being timed by hand
• Rendered brand cards and motion bookends from the client’s own logo and palette, so a second video inherits the treatment rather than rebuilding it
• Sourced every on-screen product claim to client or manufacturer documentation, and held anything unsupported as a visible approval item rather than shipping it
• Kept each job self-contained — source, config, build method, review artefacts and QC evidence in one folder — so it can be handed over or rebuilt years later

Result

• Five reels produced for one manufacturer off the same engine, each a config file rather than a production
• A new video type is a folder that imports the engine; nothing else in the studio changes
• Captions time themselves against the audio, so a script change does not mean re-timing a timeline
• Every claim on screen traceable to the document it came from, and the unsupported ones visible before release rather than after
• Covering a product line becomes an editorial decision about which products matter, not a budget decision about which ones can be afforded

Automation running here

• Product video built as a pipeline, not a shootAn instructional reel is a configuration file. Narration is synthesised, captions are aligned to the audio word by word, brand treatment and music are applied, and the finished cut assembles itself. The second video costs a fraction of the first, which is what makes covering a whole product line realistic.