Build the Versions Before You Shoot

A study on connected TV creative, commissioned by Smartly, a cross-channel advertising platform, and conducted by Forrester Consulting, puts a number on something I already believed from years on set and in the edit bay. According to the Forrester Consulting study, 95% of respondents agree that the variety and speed of creative variations have a greater impact on campaign performance than audience targeting alone. Not that targeting stops mattering. That the variety and speed of the creative itself matters more, by respondents' own account, than targeting alone.

I came up as a producer, director, editor, and brand systems designer, spending about seventeen years building and leading a creative team. From that work, I know where creative variation gets decided: in the shoot, the coverage, the edit, the approval chain upstream of any media plan. What this study describes from the media side is the exact failure I have watched happen from the production side for years.

Why does a media-side study matter to someone on the production side?

A media-side study matters to production because the study asks marketers about the output of a process I control. If creative variety and speed drive performance more than targeting alone does, the bottleneck sits upstream, in the shoot plan and the edit structure. That is not a media buying problem. It is a production planning problem, and that is the craft I have spent a career in: producer, director, editor.

The study also says many advertisers still manage creative like traditional television. In practice that means one spot, one air date, one finished cut. That description matches nearly every production I have seen asked, late, for a few more versions. The request always sounds small. It almost never is, because nothing about how the footage was shot or the edit was built anticipated a second version existing.

What actually goes wrong when the request for more versions shows up late?

Timing goes wrong. A team plans and shoots one hero cut as if it is the only deliverable, locks the edit, gets it signed off, and only then hears the campaign needs several more lengths or hooks. Every version made after that point is a salvage job, built from footage and a structure that were never designed to split apart.

That salvage work costs time in the worst possible place, after the schedule that was supposed to cover the whole job has already been spent on one cut. The editor goes hunting through footage for coverage that was never shot. The lockup gets stretched to fit a length nobody planned for. None of that is a skills gap on the team. It is a sequencing failure, and sequencing gets fixed on day one or it does not get fixed.

What has to get decided before the camera rolls?

Before the shoot, decide what has to stay fixed and what is allowed to vary across versions: how many lengths the coverage needs to support, whether the edit structure is modular enough to recut without reshooting, and whether the lockup and end card are built as a reusable system rather than artwork baked into one finished file.

On the productions I have led, a short list does most of the work:

• Which shots form the fixed core every version shares, and which are optional coverage shot only to support a shorter or longer cut
• Whether the edit timeline is built in modular sections that can be reordered or dropped, instead of one continuous cut carved apart afterward
• Whether the lockup, end card, and any required legal or brand elements exist as a system, not as artwork locked into a single export
• Who signs off on that system, so approval covers the whole version set instead of one file everyone treats as finished

Settle those before anyone calls action, and the version request that arrives later is a formality. Skip them, and that same request becomes a second production, paid for twice.

Why doesn't a new tool fix this once the edit is already locked?

A tool can speed up rendering, resizing, or delivery once the assets already exist in a form built to be split apart. It cannot invent coverage nobody shot, and it cannot retrofit a modular structure onto a cut that was built as one continuous piece. Speed only shows up where the planning already put it.

The same study found 99% of respondents want to bring social-grade creative efficacy to CTV. That number tracks, because social-native teams built the habit of variation into their format from the start: different lengths, different hooks, different cuts, planned before anything was shot. No software layer added after delivery changes the decisions that were already locked in on set and in the edit bay.

I do not take a vendor-funded number at face value just because it is dressed up as research. A study commissioned by the company selling the fix deserves the same skepticism I would give a producer who swears the schedule is fine the day before the shoot. But this particular finding matches what every production that skipped the planning step has taught me the hard way, more than once. Variation is a decision made before anyone says action. It is not a feature a platform adds after the files are already delivered.