April, 2026.- In the 2026 advertising landscape, efficiency is no longer just a goal—it’s a matter of survival. Nir Elharar, CMO of Atera, has shaken the industry by executing an AI-generated video campaign in just one week with a budget of a mere $500, a tiny fraction of the $90,000 a traditional shoot would have cost. For Elharar, the true value of this milestone lies not only in the massive cost savings—by eliminating talent, locations, and post-production line items—but in the explosion of creative freedom it enables. By not having to “protect” a monumental investment, marketing teams can afford to experiment, pivot, and generate infinite content variations for A/B testing, letting real data dictate the success of the narrative.
In this exclusive interview with Roastbrief, Nir Elharar shares a pragmatic, hype-free vision of using tools like VEO 3, Runway, and ElevenLabs. Elharar is emphatic: AI replaces the mechanical production layer but elevates the need for human judgment. For a B2B brand like Atera, where credibility is the most valuable asset, the challenge was ensuring the content didn’t feel synthetic or generic. Discover how speed became a strategic advantage to refine brand identity and why, in the age of automatic generation, quality standards must be higher than ever. This is the story of how human leadership remains in the driver’s seat of a technological engine that is redefining the cost of innovation.
1. The $500 Campaign Reality: You produced a recent AI-generated video campaign in one week for $500, compared to an estimated $90,000 for a traditional shoot. Can you walk us through where those cost savings actually came from? What line items disappeared entirely, and where did the $500 actually go?
The savings came from eliminating almost every line item that makes traditional production so expensive — crew, on-screen talent, location rental, equipment, and post-production editing. In the past, taking a campaign from concept to final cut typically took us six weeks, and every one of those days carries a cost. With AI, we compressed that to one week and kept our total spend to $500, which went entirely toward the tools we used to generate the content.
But here’s what made it even more valuable beyond the cost savings: we came out of that process with far more creative output than a traditional shoot would have ever given us. More variations, more openers, more closers — which means a much stronger ability to A/B test and let performance data tell us what’s actually working. With a traditional shoot, you protect the investment by minimizing variables. With AI, you can afford to explore.
2. What AI Replaces (And What It Doesn’t): You’ve done two campaigns entirely with generative AI, end-to-end. After this hands-on experience, what is your honest assessment of what AI genuinely replaces in the production process—and what does it absolutely not replace? Where did you still need human creativity, judgment, or craft?
In our experience, after two successful campaigns produced entirely with generative AI, the technology replaces the production layer. As I mentioned above, it cuts out the human talent needed for video production, on-screen talent, expenses needed for the location, etc. What remains irreplaceable is the human creativity that creates the video concept and refines the final product.
Instead of re-shooting if we didn’t like the cut, we could go back to the AI and tell it what we wanted it to fix. Strategic thinking, taste, and emotional judgment are assets only humans can adequately provide. AI can cut down dramatically on the labor of production; but ultimately, we still needed to decide what felt authentic, what looked too synthetic, what supported the brand, and when something simply wasn’t good enough. With AI, the options are endless. It’s up to human judgment to determine what’s meaningful, on-brand, and worth publishing.
3. The Speed vs. Quality Tradeoff: You produced this campaign in one week. In a traditional production timeline, one week might barely cover pre-production. How did the compressed timeline affect the creative process? Were there tradeoffs in quality, iteration, or creative ambition that you had to accept, or did speed become an unexpected advantage?
The one-week timeline wasn’t a forced rush; instead, it was just the result of how quickly we were able to move with this new process. Using AI fundamentally changed the shape of production entirely. To your point, one week is typically barely enough time to coordinate all the pre-production details for a traditional shoot. You have to think about coordinating people, locations, schedules, and logistics. That all takes time and money.
AI cuts all of those out. Instead, it allowed us to devote that week entirely to refining the finished product to make sure it aligned with our vision. In a lot of ways, speed actually opened up the doors to creative freedom even more since we weren’t protecting a $90,000 investment where we couldn’t afford to go in a different direction at the last minute and call everyone back to reshoot. We were able to devote our energy entirely to the creative process rather than the logistics and admin that are typically involved in shooting campaign ads.
4. The Post-Super Bowl AI Moment: This year’s Super Bowl featured a wave of AI-generated or AI-assisted ads, creating both hype and confusion. As a CMO who has actually deployed AI in production, what should marketers take away from this moment? What’s the practical, non-hypey lesson for brands considering AI for their own work?
The one-week timeline wasn’t a constraint we were working against — it was simply the natural pace of how this new process moves. With traditional production, a week barely covers pre-production: coordinating people, locations, schedules, and all the logistics that come with them. AI removes all of that friction entirely.
What that gave us was something unexpected: more creative freedom, not less. When you’re not protecting a $90,000 investment, you’re not afraid to change direction at the last minute. We could pivot, experiment, and refine without the anxiety that comes with traditional production. Our entire week was spent focused on the work itself — shaping the creative, testing ideas, and making sure the final product was right. Speed, in this case, was a creative advantage.
5. Tool Stack: VEO3, Sora, Runway, Midjourney: You used a combination of tools How did you decide which tool to use for which part of the process? Did you encounter any limitations or unexpected capabilities with these platforms, and how did you work around them?
Our primary tool was VEO 3, and we committed to it fully — pushing it as far as it could go to reach the highest production quality possible. But no single tool does everything well, and we were deliberate about filling the gaps. For camera movements that VEO 3 couldn’t deliver with the precision we needed, we brought in Runway. For image quality refinement and sharpening the final output, we used Topaz. And on the voiceover side, we used ElevenLabs to generate authentic, natural-sounding VO that didn’t feel robotic or generic.
The tool stack wasn’t chosen upfront and locked in — it evolved based on what the creative actually needed. Each tool had a specific job, and knowing when to switch between them was itself a skill our team developed through the process.
6. The Brand Voice Question: For a B2B tech brand like Atera, maintaining credibility and trust is essential. How did you ensure that an AI-generated campaign still felt authentically “Atera”—that it carried your brand voice, values, and human sensibility—rather than feeling like generic, AI-generated content?
It came down to one question we kept asking ourselves throughout the entire process: does this feel real? Not polished — real. Does it feel like something our customers would recognize as us? If the answer wasn’t an unambiguous yes, we went back and started again, regardless of how much work we’d already put in.
Everything was anchored in customer insight before a single prompt was written. We’ve spent years building Atera’s voice — our tone, our sensibility, the way we talk about the problems our customers face every day. That foundation had to carry through no matter how the content was generated. We were also deliberate about embedding human checkpoints at every stage of the process. Anything that felt exaggerated, overly glossy, or disconnected from our customers’ lived experience was rejected immediately.
The easier content becomes to produce, the higher your standards have to be. AI was our production engine. But our team was always in the driver’s seat — making sure that what we published felt authentic, resonant, and genuinely ours.
