How Czech leaders are engineering Groupon’s AI-first future

Josef Buryan is CMO at Groupon and has 15 years of experience specialising in building AI-first marketing organisations to accelerate growth. His strategic approach bridges the intersection of marketplace economics, brand strategy, and advanced AI solutions.

When you combine unfocused strategy with a technology foundation that can’t adapt quickly, you get a company that’s structurally incompatible with how fast the digital marketplace moves.

Before the new management team joined, the company was under acute financial pressure. The business had tried to be everything at once without excelling at anything. To survive, we had to stop chasing raw user volume and start focusing on reaching the right local consumers and serving their needs.

At the time, we operated across three categories: our core local service vouchers, travel deals, and physical products like electronics and apparel. The latter, our Goods vertical, faced fierce competition from global e-commerce players who simply outscaled us.

Meanwhile, the core platform had accumulated years of technical debt across multiple layers: a legacy website, a legacy app, and legacy infrastructure.

To save the business, it was time to drastically turn things around.

Redefining structure and multiplying strength

This crisis sparked a massive strategic shift, fueled by the entry of a Czech investment group and a management overhaul. To revitalise the brand, we needed to operate with the speed of a startup while leveraging the advantages of a big brand.

We committed to doing fewer things, but doing them exceptionally well. We strategically deemphasised the sale of physical products from our platform, as well as de-prioritised travel deals.

We stripped out bureaucracy and empowered people to own outcomes rather than just processes. If you want people to grow, you need to give them space to fail. At Groupon, anyone can pitch an idea, and we now have a structured pipeline to turn those concepts into live pilots.

Then, to match the accelerating pace of the market, we integrated AI. We think of AI as a multiplier of everything we do. AI agents aren’t just automating tasks; they’re providing a second pair of eyes that never gets tired and never skips a cross-check.

For example, one time an AI agent reviewing our performance reports, discovered an anomaly in wrongly attributed data. The numbers looked fine at one level of aggregation but didn’t reconcile when the agent cross-checked them at a more granular view. That discrepancy pointed to a revenue-losing issue that had slipped past our team’s standard reporting cadence. The agent caught it early, and we fixed it quickly.

Instead of mandating AI adoption, we demonstrate its potential. We show our people what they can build with their expertise, even without engineering support. The message is simple: you know your domain better than anyone, and now you have tools that let you act on that knowledge at a speed and scale that wasn’t possible before.

Capturing local intent without the manual work

Our ambition to move fast naturally extended to our paid search strategy. When searching locally, consumers use a massive, unpredictable range of search terms. No manual keyword build can cover it.

To solve this, we became early adopters of Google’s AI Max, testing it during the beta stage well before the general market rollout. AI Max brought true query intent understanding into Search, closing gaps around hyper-local queries.

We started with contained A/B testing in major verticals like Health, Beauty & Wellness, and Things To Do. The goal was to capture high-intent queries outside of our existing keyword sets.

Take a real converting query from our database: a user searching for a specific stylist in the Litchfield Park area, a small Phoenix suburb. Before AI Max, this user would likely have had a poor experience.

Under traditional keyword matching rules, the query either wouldn’t trigger our Phoenix-area salon keywords at all, or it would fail to a generic “hair salon” keyword. This would serve a national ad pointing to a category page with hundreds of unrelated deals, causing the user to bounce.

AI Max understands the query’s deep intent: the service, the stylist’s name, the neighborhood, and the state. Instead of a generic page, the user sees a relevant local ad and is more likely to convert.

Shifting from the stabilisation to growth

AI Max has enabled us to automate millions of local search queries, reducing manual effort and improving the efficiency and scalability of our marketing operations.

In 2025, we returned to growth in both billings and revenue for the first time in a decade. For our marketing team, this translates into two priorities: expanding our reach while maintaining operational discipline, and increasing Groupon’s visibility with consumers and merchants around the world.

AI is helping us advance both objectives. Through tools such as Performance Max (PMax), we can optimise campaign execution at scale and expand into new advertising surfaces more efficiently. At the same time, generative AI helps our teams create and adapt content more effectively across the breadth of our marketplace.

Our journey proves that even when standing on the brink, the right mindset, paired with the right technology, can breathe new life into the brand.