Groupon connects 42 million customers with local experiences - restaurants, events, wellness and travel - and the million-plus merchants who deliver them. We are an AI-native company in the middle of a platform transformation, moving from a deals marketplace to an experience discovery platform.

We do not run ads by hand. We build the systems that run them.

About the role

Paid Social at Groupon runs Meta across North America and 11 international markets, plus TikTok, Reddit and the channels we add next. You will lead a team of some of the best paid social specialists in the world and an army of AI agents. The agents already write the reports, pace the budgets, generate and test the creative and act on the ad accounts; the people decide what the agents should do next. Our ad accounts are tools that humans and agents call. Our creative is generated, tested and retired by systems. Our audiences come straight from our customer data platform. The question in this role is not whether to automate. It is what the best people in the loop do once the machines run the campaigns. That is yours to answer, and to build.

Who you will work with

You report to the VP Paid Media. Your peers lead SEM and PLA and Automation. You work daily with Brand and Creative, Analytics, the customer data team and Finance, and directly with Meta, TikTok and the platforms we add.

What you will be doing

• Own the channel as a P&L: spend, return, contribution margin, new customers and acquisition cost across every paid social platform. You set the targets per market and category and you build the guardrails that protect margin while the channel grows.
• Build the operating system, not the campaign. Decide which decisions stay human, and build or commission the agents and automations for everything else, with our automation team and with your own hands.
• Creative at scale: turn AI image and video generation into a testing engine, with structured tests, kill rules and a feedback loop from performance back into generation.
• Signal and audiences: own conversion signal quality, catalog quality and the sync of customer data platform audiences into every platform, so the platforms see our best customers first.
• Measurement you can defend: incrementality design, holdouts, and a single view of performance that Finance and the platforms both accept.
• New channels with a kill switch: TikTok, Reddit, Pinterest, contextual and whatever comes next. Every pilot gets a budget, a hypothesis, a date and a decision.
• Grow the team and accelerate AI capabilities: hire and develop the best specialists in the market, and raise the bar every quarter for what an AI-native paid social team can do.

What this role demands

• Six or more years in paid social, at least three of them in-house at a marketplace, e-commerce or app business where you owned the outcome and a budget of several million a year.
• You have experience with people management.
• You have a passion for AI transformation: you have tried it hands-on before, in your personal life or at work, and you are keen to learn.
• Meta depth: Advantage+ shopping, catalog and dynamic ads, Conversions API, creative testing, audience strategy. TikTok or Pinterest at scale is a plus.
• Margin and lifetime-value literacy: you can explain a spend decision to Finance in their terms and to the team in platform terms, and you know when the two disagree.
• Measurement rigor: you have run a holdout or geo test and changed a budget because of it.
• English fluent. Based in Czechia, Poland or Spain, remote.

How we operate

• Extreme Ownership. You find the problem before someone escalates it, and you own the fix, not just the investigation.
• Impact Obsessed. Platform metrics matter only when they connect to contribution margin and new customers. If ROAS is up and activations are flat, raise it.
• Speed Over Comfort. Priorities shift. The right response to uncertainty is a decision with the information available, not a request for more time.
• AI-first, for real. We encourage candidates to use AI tools during the hiring process where it adds value, and we will ask you to show how you did. The case study is graded on what you built, not on what you say about AI.