The first version was to just generate blogs or SEO content—and that’s fine, still do that. But now, if you can generate interactive, dynamic applications with minimal engineering effort, and actually build utilities people can engage with, that’s just a whole other level.
Static marketing campaigns are out; living, breathing experiences are in. More than simple storytellers, today's marketers get ahead by crafting interactive content that's less pitch, more product.
Saharsh Agrawal is the Co-Founder and CEO of YC-backed Surface Labs, a company turning websites into funnels with intelligent lead forms. He joined us to talk about why the most powerful differentiators in marketing don't have to drain time or require a full dev team.
It's a vibe: “The first version was to just generate blogs or SEO content—and that’s fine, still do that,” says Agrawal. “But now, if you can generate interactive, dynamic applications with minimal engineering effort, and actually build utilities people can engage with, that’s just a whole other level.”
With vibe marketing, startups can adopt a fast and creative approach to building tools—like the MCP directory Surface launched in just days—that deliver far more engagement than traditional content. And it's not limited to marketing. Agrawal envisions vibe marketing extending across GTM, sales, and customer success, as AI enables anyone to quickly “think of something and manifest it.”
Finding alpha: Sinking days, weeks, or months into a project doesn't inherently make it great. In fact, it might be a hindrance. “Marketers don’t necessarily need to go and buy large platforms or point solutions,” he says. “You can code up a basic version in a couple hours, if not less.”
That speed unlocks what he calls go-to-market alpha: discovering creative, high-leverage tactics that outperform by getting to market before others catch on. “How do you find the things that help you win before the rest of the market catches up?” Agrawal asks. “If marketers can do that on their own—without waiting for an engineer—that’s how they accelerate, test ideas faster, and find alpha.”
Trying to be AI native for the sake of it is silly. It should be because that's the best way to run your company, not because you have an ideological belief that you need to inject AI into every crevice.
AI, ASAP: With AI tools abound, utility still reigns. "Trying to be AI native for the sake of it is silly," says Agrawal. "It should be because that's the best way to run your company, not because you have an ideological belief that you need to inject AI into every crevice."
But with the right tools, early adoption is crucial, and not just for immediate gains. It allows teams to train and prepare for when AI systems become more advanced, avoiding the struggle against steep learning curves. "You should start using it now, even if it's not perfect," Agrawal advises.
The data moat: As tools to build applications get faster and easier, Agrawal argues that data—not code—is the real advantage. “The only real moat is owning the best data,” says Agrawal. “It’s become so easy on the application layer to build workflows and integrations. But if your system is where the best data is produced, you win.”
Basic A/B testing isn’t enough anymore. “You need to go a lot deeper,” he adds. That’s why he’s watching developments like OpenAI’s one-click shopping and memory features closely. “They’re laying the foundation for a consumer-level data layer you can build new experiences on top of,” Agrawal says. “That’s going to unlock a lot of new monetization channels.”
