Data Scientist, Marketing

The Data team at SpaceXSI is a small team that works across the company to inform business and product decisions, and we’re hiring the first member of our team to focus fully on marketing. Performance Marketing at SpaceXSI drives scaled acquisition across 10+ channels for Grok, Grok Bot, and our full portfolio of AI products, across consumer, SMB, startup, and enterprise audiences.

This role will sit in the central Data team and partner directly with Marketing leadership to own the measurement and analysis of performance marketing, growth marketing programs, and overall marketing measurement. You'll use your expertise to inform where the next dollar goes as we scale paid acquisition, while building the foundations of a world class marketing data team. Your job is to help us understand the impact of our activities and what we should do next.

What you'll do

Build and own the measurement layer for performance marketing: economics and performance by channel, campaign, audience, geo, and creative, across both PLG and Sales-Led revenue lines

Design incrementality tests (geo, audience, and conversion lift) with channel owners, and use the results to move budget and calibrate attribution, not just to fill a slide

Stand up and maintain our media-mix model, and pair it with test results to answer the big allocation questions: which channels to scale, which to cut, and where we're hitting saturation

Own the quality of the predictive LTV signal we send to Meta, Google, and the other networks: validate it against realized revenue, catch drift and breakage early, and partner with Data and Engineering to improve it

Keep pipeline source and campaign attribution separate and clean, so self-reported source, promo codes, click-based attribution, and CRM pipeline each answer their own question without double counting revenue

You may be a fit if

You have 7+ years of experience in marketing analytics, marketing science, or data science supporting paid acquisition

You have strong SQL and python skills and are comfortable building both data pipelines and predictive models

You've run hold-out or incrementality tests and worked with MMM and attribution models in production, and you know the limitations of each

You understand how ad platforms report, optimize, and ingest conversion signals, and you don't take platform-reported numbers at face value

You communicate clearly with channel managers, Finance, and leadership, and you're honest about what the data does and doesn't support

You're self-directed, comfortable owning ambiguous problems end to end