How to Scope a Video System Before You Build It

Published by Videonomy — Video Production Systems, Built to Scale

---

The most common reason a video automation project fails isn’t technical. It’s a scoping problem — the wrong question was answered, or the right question was never asked.

Before any pipeline gets built, there’s a conversation that determines everything downstream. Getting it right takes about an hour. Getting it wrong can cost months.

Here’s how we approach it.

Start with the Repeating Unit

The first question we ask isn’t “what video do you want?” It’s “what is the smallest unit of content you produce repeatedly?”

For a fashion brand, that’s a product. For a SaaS company, it’s a feature demo. For a sports club, it’s a personalised message per member. For an event company, it’s a speaker graphic.

That repeating unit is the atom of your system. Everything else — the templates, the data connections, the rendering infrastructure — is built to serve it. If you can’t define it clearly, the rest of the scoping conversation won’t work.

The test: can you write the unit down in one noun? Product. Member. Speaker. Course. Donor.

If you need a paragraph, you’re describing a category, not a unit. Narrow it further.

Map the Data

Once the unit is defined, the next question is: what data does a single instance need?

Write it out like a variable list. For a product video:

• product_name
• product_image
• price
• category
• short_description

Now: where does each variable live? In a Shopify catalogue, a Google Sheet, an Airtable base, a Notion database, a CRM?

This mapping exercise usually surfaces two things: data that’s already well-structured (good), and data that lives in inconsistent formats across multiple places (a scoping issue that needs resolving before any automation can work reliably).

Don’t try to scope the automation around broken data. Fix the data first, or accept the constraint and scope around it.

Define the Output Format

Not just “video” — be specific.

• Duration: 15 seconds, 30 seconds, 90 seconds?
• Aspect ratio: 16:9 for YouTube, 9:16 for Stories/Reels, 1:1 for feed?
• Does it need subtitles? A voiceover? Motion text only?
• Where does it get delivered — downloaded as MP4, posted directly to a channel, emailed to a recipient?

Each of these choices affects the render engine, the template architecture, and the delivery mechanism. They’re not cosmetic decisions. Define them upfront.

Understand the Volume and Cadence

A system that produces 10 videos a month is a different infrastructure investment than one that produces 10,000. Not necessarily more expensive to build, but different in where the constraints live.

Ask:

• How many videos per month at steady state?
• Are there spike periods (product launches, seasonal sales, annual events)?
• Is production reactive (triggered when something happens) or batch (scheduled runs)?

Volume also tells you how much time needs to be saved to make the system economically justified. If you’re producing 20 videos a month at 3 hours each, that’s 60 hours. If a system can reduce that to 10 minutes each, that’s 56.7 hours recovered. The build cost pays itself back in a measurable timeframe.

Decide Who Operates It

A system designed to be operated by a technical team looks different from one designed for a marketing assistant.

The more complex the operator permissions, the more the interface matters. A dashboard-first approach (a simple Airtable view, a Notion form, a branded web interface) is almost always the right call if non-technical staff will be triggering production runs.

Ask: who will press the button? And what level of confidence do they need before they’re comfortable pressing it?

The Scoping Output

By the end of this exercise, you should have:

1. A defined repeating unit
2. A variable list with sources
3. An output spec (duration, ratio, format, delivery)
4. Volume and cadence figures
5. A named operator and interface preference

That’s not a brief. That’s a scoping document — the foundation everything else is built on. With it, a system can be architected, costed, and delivered. Without it, you’re building something, but not necessarily the thing you need.

---

If you’re at this stage and want a second opinion on your scoping, let’s talk. We review video automation briefs regularly and can usually identify the key decisions in one conversation.