From Swifties to seat sales: How AI is revolutionizing SeatGeek’s marketing

‍When you've got thousands of events and millions of potential customers, how do you match the right ticket to the right fan at exactly the right moment? 

In this candid conversation, Jason Davis, Founder and CEO of Simon Data, sits down with Steve Mastrocola, Senior Director of Audience Marketing at SeatGeek, to unpack how AI is transforming audience targeting, eliminating marketing guesswork, and helping fans never miss the events they'll love. Note: This interview has been edited for clarity and length. You can watch the full interview below.

Steve, how do you tackle customer segmentation at SeatGeek? What headaches are you dealing with on both the strategy and execution sides?

Steve: Well, like many e-commerce companies, we face unique challenges with lifecycle marketing. But we also have the added dynamic of what's happening in the live event landscape — and how to personalize our marketing for customers.

Our biggest challenge right now is figuring out which events to promote to which person at the right time, with so many events to choose from and so many dynamics going on in the live event industry. For example, if Taylor Swift is touring and you're not a huge Swift fan, it doesn’t make sense for us to promote her tour to you.

Jason: I love the Taylor Swift example. When she is touring, the audiences you want to market to aren't always obvious. If I'm a Taylor Swift fan in New York, you'll probably promote her show at Madison Square Garden. But many people will travel to see Taylor Swift as well. Who are those people, and how do you think about them?

Steve: There are so many aspects to consider. How do we know if you like Taylor Swift? Maybe you've shown interest in her or in similar artists. But how do we determine which artists are similar? Where is she playing? Where do you live? Are you willing to travel? How far?

We have access to all this customer data, but my team isn't composed of data scientists. We use tools like Simon Data to generate audiences and build them at scale, but our challenge is in managing all these data points effectively without someone having to make arbitrary decisions or write code that might not be scalable or transparent.

We've been talking about what we call “100x data” – the opportunity to unlock that next generation of data. How are you thinking about leveraging your data to the fullest?

Steve: Much of our prioritization today is dictated by scale and popularity. So, NFL, MLB, NFL, or major concert artists where there's significant audience interest. But one challenge is that there are thousands of smaller artists and sports teams that collectively add up to a lot of potential revenue. Managing all of that simultaneously is much more difficult than managing just the NFL.

We also face daily decisions like: "We have an NFL campaign going out today, but Beyoncé tickets are going on sale. How do we prioritize?" These decisions are often made manually right now, and we'd love more automated, machine-driven processes for both operational efficiency and greater accuracy. Essentially, how do we remove human biases in favor of data-driven decisions?

You mentioned that AI reduces the need for humans to make arbitrary data choices. What do you mean by that?

Steve: When we choose to promote a music artist, we're using our best judgment based on what worked at scale in the past. But as we get more data, you lose economies of scale in how much you can manually evaluate. 

People also bring their own biases, such as, "This worked before, so I'll use it again," or, "I personally like these two artists, so they must be similar," when that might not be true for everyone.

By having machines make these decisions using the best possible data, it frees my team to focus more on strategy: when to send something, what creative aspects to test, and frequency of messages. It reduces the burden of making potentially poor decisions while opening operational doors to focus on areas where human creativity and judgment are more valuable.

Speaking of data points, we just built this cool weather agent into our platform. How might something like weather data actually help your marketing?

Steve: Weather plays a huge role in event attendance! For baseball events, a 74-degree sunny day is much more appealing than a rainy one. Arena events have higher attendance when the weather's bad because people want to be inside.

Having access to weather data helps us improve our marketing messages. A push notification saying, "It's going to be 75 and sunny tomorrow – perfect for a day at Citi Field!" is more compelling than a generic message.

The power of third-party platforms serving this data broadly is efficiency. If I wanted to leverage weather data internally, I'd need several teams at SeatGeek to do a bunch of work. But weather is universal, and if a platform like Simon Data brings that capability to the table, it saves me advocating internally and allows all clients to leverage it at once.

As you know, we recently launched Simon AI. But our approach to AI is different: Marketers start by inputting their goal and letting the AI find the right audience to achieve it. How do you see that changing the way your team works?

Steve: Currently, we think: "We have this campaign, how do we find an audience?" But it should be flipped to start with a goal: "We want to sell last-minute baseball tickets", and then having an AI agent identify people who buy last-minute MLB tickets.

The AI could even define what "last-minute" means for different people. For some, it might be buying an hour before the game. For me, with children, I can't make that decision in an hour – it might be 3-4 days out. Unlocking these insights at scale lets us market to someone based on how they view "last-minute" versus how we view “last-minute.”

Let's zoom out a bit. What advice would you give to marketing leaders who are hesitant about jumping on the AI bandwagon?

Steve: The biggest thing is keeping an open mind. I've been at SeatGeek for eight years, and in my tenure, I've been involved with CRM in some capacity. I could have an ego and say, "I know what works," but being here that long  doing the same thing is also a detriment, right? It means I might have wrong biases as things have changed.

So, having that open-mindedness that machines can calculate and do certain tasks better is incredibly helpful. Stress to your team the benefits of AI – it's not replacing anybody; it's adding bandwidth so you can focus on more strategic, career-rewarding aspects versus execution and operations.

Also, think about how you use AI in your daily life and how it could apply to your work. We've seen many people start with copy testing by having AI generate subject lines. That was basic but helpful, but AI goes much further. Creating audiences and effective targeting are far more complicated, take more bandwidth, and often have more impact on your company.

My brother got married last year, and he literally wrote his vows by going to ChatGPT. They were good! We would never've known that he did it that way. The breadth and quality of what's out there are pretty cool.

Fast forward a year from now. What does success look like? How will you know if all this AI stuff is actually working?

Steve: Success looks like eliminating those constant decisions and back-and-forth discussions my team has around, "Should this campaign go today or tomorrow?" or, "Who should we include here?"

When we first started CRM at SeatGeek, we literally had a Google calendar for campaign scheduling. That evolved into trigger-based campaigns and recurring newsletters, but we still debate things like: "It's Selection Sunday for March Madness, but there's also a concert on sale that day – which should we prioritize?"

Success is those questions just going away. We're not having a person make decisions based on what we think is best from past experience. Instead, AI is actually crunching the data, and it's not an either/or question. The machines decide which people go where, maximizing response and revenue while helping our users get the best experience and recommendations.

Jason: It is fascinating how efficiency and operating leverage drive better results for both the bottom and top lines. When I think about AI's potential, it's allowing your team to do more in a more informed way than today while using it to drive personalization that makes everything they do smarter.

Steve: Absolutely — especially in the CRM world where people get so many emails and push notifications, and it's getting easier to opt out. Having quality messages in the right amounts to the right people is more important than ever. It becomes more of a one-to-one relationship that builds trust and prevents unsubscribes.