Growth Marketer

Sunset•New York, NY

4d•$175,000 - $235,000•Onsite

About The Position

Sunset was founded to help founders, initially by supporting startups through shutting down, and has since expanded to unlocking new revenue streams for businesses. In 2025, Sunset identified that the data generated daily by companies (collaboration, communication, building) is valuable training data for AI models. Sunset partners with frontier AI labs, providing them with this proprietary data. The company has experienced rapid growth, scaling from $0 to a $XXM run rate in months, and has secured funding from top-tier investors. Joining Sunset now offers the opportunity to take on significant responsibility, grow with the company, partner with leading companies, and be part of building a category-defining business from the ground floor.

Requirements

Excited about building in-person, 5 days a week, from our office in Dumbo
Want to work hard and strive for world-class levels in performance
3-5+ years in growth or performance marketing, ideally somewhere early-stage where you owned channels end-to-end
Thrive in ambiguity: no fixed playbook exists yet, and you're comfortable making fast decisions with incomplete data
Creative and technical self-sufficiency: you build real things yourself (landing pages, workflows, automations) with AI tools, rather than briefing someone else
Genuine experimenter: breadth over narrow-channel depth, always testing something new
Data-literate: can read your own results well enough to know what's actually working

Nice To Haves

Experience marketing to founders/startups or other B2B/technical buyers
Familiarity with AI training, data licensing, or model evaluation
Experience launching a channel from zero, with no existing playbook to follow

Responsibilities

Own top-of-funnel growth end-to-end: pick channels, run experiments, and scale what works
Generate qualified marketing leads (MQLs) for both our Wind-down and Data Partnerships motions
Build and ship your own landing pages, campaigns, and workflows using AI tools, rather than waiting on a dedicated design or engineering team
Make fast calls with incomplete data: kill what's not working quickly, double down on what is
Track and report your own results honestly, since experimentation is worthless without real measurement