GPT-3 turned Waymark’s DIY video maker into a scriptwriter that works

For years, Waymark had cracked one half of a very hard problem. Their platform let anyone create sleek, broadcast-ready video ads without technical chops. But there was a catch—and it was a big one. The company’s own research revealed a brutal statistic: two out of three users were getting stuck before they even started, paralyzed by the blank page and the simple question of what to say. Their existing fix, a rules-based script generator, churned out copy so generic it often required heavy editing or, worse, led users to abandon the project entirely.

Founder Nathan Labenz had spent half a decade hunting for a solution, kicking the tires on every AI writing tool that surfaced. Nothing stuck. “I’ve tried every AI-powered product available over the last five years, but found nothing that could effectively summarize a business’s online footprint, let alone write effective marketing copy, until GPT-3,” Labenz said. That changed when they began fine-tuning models on GPT-3. The integration didn’t just tweak the old feature—it fundamentally altered the product’s DNA, shifting it from a do-it-yourself editing suite into what the team now calls a natural-language video creation platform. Users input their business details and get back an original, custom script in seconds.

The impact was immediate and tangible. CEO Alex Persky-Stern noted that customers are now spending far less time in the editing weeds because the first draft is actually usable. The copy is sharper, more relevant, and tailored to the specific business rather than a broad industry vertical. For agencies and platforms managing ads for hundreds of local businesses—think franchise networks or multi-location retailers—this isn’t just a convenience. It eliminates a major bottleneck, letting them pitch and produce for more clients without scaling their headcount.

What’s interesting here isn’t just that AI can write copy. It’s that Waymark’s team had already tried and failed with off-the-shelf AI for years. The leap from a frustrating template system to a model that synthesizes a business’s entire online presence into a coherent script points to a future where the “creative block” is less of a human failing and more of a solvable software problem. The question is whether this done-for-you model becomes the baseline expectation for creative tools, and what that means for the copywriters who built their careers on solving this exact pain point.