In Q4 2024, an early-access team needed 9 days to move from brief to first cut. By the next quarter, a generative workflow with no camera crew reduced that cycle to 7 hours. They produced 4 ad variants in that session. The two-person team was a copywriter and marketing manager, neither with prior video editing experience. This outcome is not universal across every project type, but shows what is consistently achievable for a defined class of ad productions.
Where production time goes
Anyone who has run a traditional ad production knows the calendar is largely waiting. First comes the director's creative treatment, then location scout confirmation, then conversion of the talent hold into a confirmed booking. After that, there is the first cut, client review revisions, the revised cut, and final approval.
For a modest 30-second brand awareness ad, pre-production coordination takes roughly two days: shot list, casting, location, scheduling, and brief alignment with the production company. The shoot takes one day. Post-production revisions consume three to four days. Across that span, the hands-on technical work of capturing and cutting footage may total only 8 hours. The rest is coordination overhead and the serial nature of review cycles.
Generative video still needs creative direction. It removes the logistical chain that makes revisions costly. When the model creates a new cut, you do not wait for an editor's next open calendar slot. You wait for the generation job, which completes in minutes on our priority queue.
Three stages, compressed
Pre-production sees the greatest compression. In a generative workflow, the inputs that remain are the script and selected reference frames. Preparing them for a straightforward 30-second ad takes roughly half a day. Shot lists, casting calls, location permits, production briefs, talent contracts, and equipment hire disappear because they assumed a camera would be on set. Without one, the pre-production list shrinks by around 80 percent.
Production becomes a generation session. Creating 4 to 6 variant cuts of a 30-second ad with the same script and minor parameter changes typically takes under 3 hours, including output review and selecting the direction to develop. Decisions still matter, but they are creative choices rather than logistical execution.
Revision cycles remain, but their nature changes. Rather than send notes to an editor and wait until the next day, you generate again with updated inputs. The iteration loop moves from 24 hours to under an hour. For a team that usually completes 3 client revision rounds, that removes 2 to 3 days from the schedule by itself.
The human work that remains
The script carries more weight in a generative workflow than in traditional production. With a camera crew, a skilled director can make real-time creative calls on set when the brief is thin. With a generative model, there is no director on set. It interprets the written direction, so vague scripting creates vague output, and parameter tuning cannot recover a weak brief.
Choosing reference frames also demands real visual judgment. Uploaded frames anchor the output to a particular lighting environment, color palette, and motion energy. Teams in our early-access cohort that rushed this step repeatedly said the result felt flat or generic. That points to a brief limitation, not a model limitation, much like arriving at a shoot without a mood board and expecting the DP to infer the aesthetic.
Client approval cycles stay the same. The model can create 6 variants while a traditional editor creates one, but a person with sign-off authority still has to select the direction. We have seen teams make strong first-generation cuts, then spend 3 days in internal review because decisions lagged behind production. Generative video cannot repair a slow approval workflow.
Pilot data findings
In our 2025 early-access cohort, the median brief-to-first-cut time fell from approximately 4 days to under 12 hours for brand awareness and campaign ad projects in the 15-to-60-second range. The early-access pilot data covered teams producing content series and campaign launches, not single-off productions with external approval chains. The largest compression appeared when the marketing team owned both the brief and the internal approval process.
For single hero videos with complex external client approvals, the reduction was smaller, from 7-9 days to roughly 2-3 days. Generation was fast, but the organization around it was not. The distinction matters: the model speeds up production phases, not people phases.
When traditional production wins
Generative video does not suit every brief. When you need a specific person on screen, a particular physical location with recognizable details, or documentary footage of a real event, a production crew remains the faster, more reliable route. Our models currently do not create photorealistic human performances at the quality needed for certain brand categories, including luxury goods, healthcare, and financial services, where viewers need confidence that the person or place is real.
The greatest compression occurs with category video: brand awareness campaigns, product launch announcements, social content series, and brand films conveying energy and values instead of a specific product in physical space. This category takes a significant share of a mid-size brand's yearly video budget, making the timeline case especially compelling.
Using recovered time
The more useful question is what happens with the time recovered, not only how quickly production moves. In our pilot cohort, the teams gaining the most from the shorter timeline used it for more format variants: 16:9 for connected TV, 9:16 for vertical social placements, and 1:1 for square mid-feed, all from one generation session with minor output adjustments. They did not simply make one ad faster. They created a complete placement set in the time that once produced one format.
Moving from one format to many in a single session is what changes a small team's competitive position. The gain is not just speed, but consistent creative across every channel without the per-format production cost that made multi-format campaigns impractical for teams under 10 people. An ad that once needed three separate production days for three formats now takes one morning.
