Video Production
Video production in marketing covers everything from a thirty-second paid ad to long-form anchor content: scripting, shooting, editing, and versioning for each placement. It became a priority line item because video dominates the major platforms and because agencies sell production value as the product. Craft is evaluated after the message is proven. The first question is whether the message in the script is one the market has already shown it responds to.
How it actually works
The standard engagement runs brief to agency to hero asset. The team writes a creative brief, the agency produces a polished flagship video, and the asset ships across channels with success measured in views, completion rate, and recall. The stated goal is usually awareness.
Two premises in that engagement do not hold up. The first treats branding and selling as separate jobs, one for the polished video and one for the performance ads. Content should always have a financial objective, and a goal stated as awareness has no return attached to it, which makes it impossible to evaluate. The second holds that audiences only watch short video. Nearly all of the most-watched creators on the major platforms win with long-form, a pattern we document in [the attention span fallacy](/insights/why-your-vendors-win#the-attention-span-fallacy). Viewers stay with content that addresses what they already want.
When the message is not grounded in market data, spend concentrates on polish. The expensive asset ships untested because of the amount already invested, and the platforms' view metrics measure watchability, not purchase intent. The order of operations corrects this: prove the message in small, inexpensive tests first, and reserve the production investment for concepts that have already sold. The arithmetic favors the test whenever its cost is below the production cost multiplied by the probability the concept fails, and because production budgets and the range of possible outcomes both grow with scale, the case for testing first gets stronger as the planned shoot gets bigger.
In practice
One asset shows what a proven message returns on production spend. One of our longest-running ads ran over 3 minutes with a watch retention of 80 to 90% for almost 3 years. It required one edit, removing a single word after an algorithm update, and then ran for years after that. That ad took a business from $300k per year to $20M per year and a healthy exit. The case is documented in [the attention span fallacy](/insights/why-your-vendors-win#the-attention-span-fallacy).
Where we come in
We calibrate creative to the customer's stage of awareness, and a claim not backed by data within the target market does not end up in scripts or ad creative. We test small and prove concepts before investing in anchor content, an approach applied across engagements with MrCool, Potbelly, Skull Shaver, Transparent Labs, and Old School Labs, where revenue grew 87% in ten months.
Start a Revenue Health Pre-Assessment →See it in action
Related terms
- Anchor content
- The flagship long-form asset a campaign is built around, with shorter cuts and placements derived from it. It is the largest creative investment, which is why it gets tested last, after the message has proven out small.
- Creative testing
- Running low-cost variants of a message or concept against real audiences before committing production budget. The test establishes evidence; the production buys reach.
- Stages of awareness
- Eugene Schwartz's five stages, from unaware to most aware. Each stage needs different creative, which is why one video rarely works across the whole journey.
- Direct response
- Creative built to produce a measurable action now: a click, a lead, a sale. The tradition behind David Ogilvy's "We sell or else" (Ogilvy on Advertising, 1983).
- Completion rate
- The share of viewers who watch to the end. It measures whether the video holds attention, not whether it moves anyone toward a purchase.
- Hook
- The opening seconds that determine whether a viewer stays. On paid placements it is the highest-leverage part of the asset, and the part most worth testing in variants.