Native wants to choose the marketing idea before generating the ad

Native is tackling a decision that comes before an AI marketing tool writes a caption or makes an image: which idea is worth producing? Its newly announced Norn system ranks concepts using a business’s brand, past social performance and feedback, then feeds selected ideas into the company’s content-generation and publishing workflow.

The startup calls that selection process artificial intuition. In practical terms, it is an attempt to make software do more of the editorial choosing that still falls to a business owner after a model produces a long list of possibilities.

Native’s launch describes a workflow that begins with a company’s website. The system builds context about the business, proposes angles and formats, and ranks ideas before generating the finished content. Frontier AI models handle production, while Norn is intended to guide what they should make.

That separation is the interesting part of the announcement. Generating more material is useful only if someone can decide what fits the business and what deserves a place in its publishing calendar. Native is trying to put that decision inside the same product that makes and distributes the work.

Watch Native’s launch demonstration.

The company says Norn learns from the content users accept, rewrite or reject, alongside the performance of published posts. Those signals serve different purposes: an owner’s rejection can reveal a mismatch with the brand even when a suggested idea might attract attention. Performance feedback, meanwhile, can help assess what happened once a piece reached an audience.

After review, Native can schedule or publish content and bring results back into the workflow. Its product covers social posts and advertisements as well as image and video content. The intended customer is a business that needs regular marketing without turning every post into a separate production project.

Native reports serving more than 2,000 businesses and reaching $2 million in annual recurring revenue within its first six months. Those are company-reported figures from the launch announcement, not independently audited results. The founders also say they use the product for Native’s own marketing, giving them a direct reason to notice where its output or judgment falls short.

The team combines content creation, marketing operations and machine learning. Benjamin Bekken’s public background includes years making online content and a stint as a Minecraft YouTuber. Håkon Græsberg brings machine-learning experience and previously served as Wepost’s CTO. Trond Hauga’s experience includes running an agency and a music festival. Wepost merged with Native in January 2026, according to the company’s public YC profile.

Native is part of Y Combinator’s Fall 2026 batch. Its announcement also points toward making Norn available beyond its own application through an API. That offering is still in development with a waitlist; it should not be confused with an already available developer service.

The launch does not establish that ranking ideas improves results for every business or campaign. That will depend on the quality of the feedback and on whether the system distinguishes a useful commercial outcome from a post that simply attracts engagement. For customers, the immediate test is more concrete: does Native’s shortlist reduce the amount of choosing and rewriting needed before a post can go out?

‍