We used to shoot all our social ad clips in-house. Switched to Sora to cut costs and increase output. It’s been two months. Some wins, but major limitations are blocking us.
Pros:
* Speed and volume are unmatched. We can generate dozens of concepts in an hour.
* Cost per asset is far lower than a production day.
* Good for abstract visuals or backgrounds where realism isn't critical.
Cons / blockers:
* **Product consistency is impossible.** We sell a physical product. Sora cannot generate the same product from multiple angles with perfect consistency. Colors, logos, and minor details shift. This is a dealbreaker for true product-focused ads.
* **Integration into our existing asset pipeline is clunky.** We have a brand kit, specific fonts, and existing footage. Sora outputs feel "separate." We spend more time compositing and correcting than expected.
* **No real control over typography or on-screen text.** You can *describe* text, but the styling and placement is a gamble. We have to add all text in post, which kills the speed advantage for many projects.
Question for others using AI video for performance ads:
How are you handling brand/product consistency? Are you only using Sora for certain ad types (like awareness top-funnel) and not for direct response? What's your post-production workflow look like?
null