Head of SEO & AI Search in London

Groupon connects 42 million customers with local experiences - restaurants, events, wellness and travel - and the million-plus merchants who deliver them. Our mission is to get people offline and into the real world at prices that make it possible. We're an AI-native company in the middle of a platform transformation, moving from a deals marketplace to an experience discovery platform that works for customers and merchants at the same time.

AI Overviews now appear on the large majority of the local queries we track in North America. ChatGPT, Claude and Gemini recommend places by name, and agents are starting to search, compare and transact without ever loading a page. Groupon's Local estate runs to a six-figure number of live indexable pages in North America after pruning, sitting inside an addressable space of roughly 1,000 cities and 50+ categories, and only a minority of them earn. The wider domain is bigger than Local alone, and the role is global, the majority of organic revenue is North America, the rest is the UK, Germany, France, Spain, Italy and Australia. Organic revenue is growing year on year and the demand, presence and capture instrumentation already runs, so this is neither a rescue case nor a discovery quarter. The prune is largely behind us.

What is missing is the next decision, city by city and category by category: which local pages Groupon funds, which it rebuilds, which it consolidates and which it stops publishing. Structure, craft content and technical hygiene serve that decision. You own it as a hands-on operator and a sharp analyst, mostly operator, and the next twelve months decide whether these pages become the structure users, Google and agents reach for.

Who you'll work with

You will report to Josef Buryan, Chief Marketing Officer, who owns Marketing end to end. You lead an SEO team of four to five (AI content, technical SEO, SEO data science, core SEO) plus a shared engineering squad on the web platform working from your backlog. You partner weekly with Product, Engineering, Data, Paid Search and Brand, and you hold Supply to inventory density per city and category with the evidence to back it. You report monthly to the CEO and CFO in the "battlefield view" - a battlefield being one city-and-category market on the domain, sized by its demand, our presence, the comparable competitor's presence, a realistic target and the conversion that decides whether investment pays. This seat is visible at the top of the company from the first month, and the CEO will challenge a number on your page. Expect a shared eng squad and no incremental headcount, weekly 5/15s in text not decks, and cross-functional teams that need structure and pushing to move. Diagnosis without shipped work does not survive here.

What you'll be doing

• Deciding the Local estate: which cities get coverage, which categories get a dedicated page, which of the four Local page types (city root, city things-to-do, category-in-city, deal) owns each intent, what supply threshold makes a page viable, and what gets rebuilt, consolidated or removed - with the volume affected, the expected click loss, the expected upside and the revenue effect stated up front.
• Designing the GEO x Category architecture and internal linking structure that carries that decision, so the city to category to deal path is explicit, reachable within three to four clicks, with no orphaned pages and no diluted authority - first version of the structure and the build order inside two weeks, then shipped against it.
• Owning the SEO engineering backlog as business owner so the funded pages actually get rebuilt: writing acceptance criteria, holding backlog reviews, blocking releases that break the standards you set - a page-weight budget on transferred bytes before interaction, an uncompressed-HTML gate at deploy under the 2 MB Googlebot limit, one media player click-to-play, response time - and escalating when platform ships without SEO sign-off.
• Governing third-party scripts on every page type (inventory, owner, purpose, keep or kill) and owning crawler and rendering policy across all markets, including AI crawler policy as a conscious decision rather than a default.
• Running the content factory that rebuilds them, as a GenAI production system - structured inputs, research, reusable skills, human checkpoints, QA - behind a page template, an editorial line and a quality gate; stopping publish until the craft bar exists and fixing or removing what does not meet it.
• Building content so it travels: pages and assets that Display, SEM, Influencers and Social can reuse. One production cost, four channels.
• Driving GEO and AI search at city and category level so the funded pages are the ones that get cited: AI Overviews presence, machine-discovery endpoints, structured deal data, connectors to Claude, ChatGPT and Gemini, and Reddit and LinkedIn where they move citations - measured geolocated rather than nationally.
• Reading the battlefield instrumentation that already exists, knowing where it is strong and where it is thin, and deciding from it; rebuilding it is not the first project.
• Leading the team and the shared eng partnership day to day: hire and exit, hold reports to weekly output, keep the backlog full and specified, and produce the monthly battlefield MBR the CEO and CFO can act on.
• Defaulting to AI agents for audits, content production, monitoring and reporting, and showing the ROI of what you built versus what you consume.

What success looks like

• Day 7: You have oriented on the battlefield view, the current index and craft gaps, and the eng backlog. You have set direction from the data at hand and written the first tickets, not a research plan for month three.
• Day 14: GEO x Category and internal linking one-pager delivered: what exists, the main hubs, what is missing, the build order. Plus the first Local allocation call - the cities and categories funded first, and the first to stop. First acceptance criteria for platform hygiene (page weight, HTML size, scripts) drafted for engineering.
• Day 21: Content template, editorial line and production skill live; first pages published to the craft bar or bad pages stopped. You are running the weekly 5/15 to standard and holding eng review without hand-holding on ticket quality.
• Day 60: Team assessment complete; open SEO role hired or in final stage; every report's weekly update to standard. First battlefield pages live for at least one city and one category against a target, with a scale plan. You own the monthly MBR shape in the battlefield view without rewriting it from scratch each time.

What this role demands

• You can demonstrably state an index and investment policy for a six-figure Local estate - what stays, what goes, what gets rebuilt - and defend it with numbers, assumptions and the first ticket you would write.
• Ship from incomplete data on a messy programmatic or marketplace site: name brief-to-live work with dates and numbers moved, not a six-month framework before the first ticket.
• Brief engineers without a translator: write acceptance criteria, hold backlog reviews, and escalate when page weight, HTML size against the 2 MB Googlebot limit, scripts or crawler policy break the standard.
• Own SEO on a large multi-market website, at the order of 500k+ programmatic pages - that is the scale you have operated at, and our domain is larger than the Local estate alone - including marketplace or local-commerce dynamics: supply density per city and category, local pack behaviour, cannibalisation between near-identical pages, prune and consolidate.
• Run GEO and AI search hands-on at city and category level with a measurement method, not in general terms and not from a conference stage.
• Measure Local honestly: geolocated SERP measurement rather than national rank tracking, awareness that search console aggregation hides city-level divergence and that page-type aggregation masks cannibalisation.
• Design and evaluate a GenAI content production system: structured inputs, research, reusable skills, human checkpoints, QA.
• Operate AI-native day to day: walk through your tools, agents and skills, what you personally run every week, and what it replaced.
• Hire, develop and exit SEO talent, and hold a shared engineering partnership as the business owner of the backlog.
• Absorb a crawl audit or a battlefield dataset, find the three things that matter and state a decision in minutes, explaining why rather than what.
• Write a one-page business view (size, capture, competitor, target, three decisions) in English text on the day asked, hold the structure when one number is challenged, and keep pushing cross-functional teams when they stall.

How we operate

We hire for how people work, not just what they can do. The Operating Principles most relevant to this role:

• Extreme Ownership: You own the Local index decision, page weight, scripts, crawler policy and content quality as the business owner even when engineering, product or supply made the change. "They shipped without telling us" is an escalation failure, not a status update.
• Speed Over Comfort: A defensible v1 in days, then iteration. You do not hold the team for certainty, and "meaningful results in six months" is the wrong rhythm for this seat.
• Impact Obsessed: You prioritise by revenue at stake times winnability per battlefield, report in revenue and clicks not rankings, and kill low-value work.