“I was so drawn to that data storytelling piece. You can build a model and do analytics, but it will never get implemented if you can’t present it correctly. And you’re never going to drive value from it. So being able to do both is an essential balancing act.”
Meet Morgan Ferguson, currently a Lead of Data Strategy at The Walt Disney Company and an IAA ’22 alumna. Morgan started her journey at Disney as a Senior Data Scientist, where her work wasn’t focused on Disney’s media empire, but on something equally captivating – the Disney Experiences segment, which includes Theme Parks, Resorts, and Cruises. With a background in technology consulting, she brings a blend of technical expertise and creative problem-solving to enhance Guest experiences and optimize operations across Disney’s physical destinations. In this interview, Morgan shares some ways how Disney leverages data science to enhance Guest interactions and how data informs decision-making while balancing creativity and analytics.
Glynn Smith: “For years Disney has been a master at perfecting the customer experience with the goal of making every experience magical. What are some of the ways that Disney uses data science?”
Morgan Ferguson: “That’s a good question! Recently, there’s been a significant initiative to ensure that every commercial decision has explored the impacts on the customer experience. So, suppose we’re thinking about raising ticket prices or opening a new land. In that case, we want to see how that will impact our customer experience and make sure that we’re making informed decisions and that we can balance customer satisfaction with the company’s goal of making a profit.
There are hundreds of analytics and data science teams within the company. Every team is doing things a little bit differently. And every team works on different data. So, it varies from team to team and from role to role, but I’m definitely happy to share my experience.
We get our customer experience data from lots of different places. There’s actually an entire team dedicated to consumer insights and marketing analytics called ‘CIMA.’ They do a lot of research on multiple audiences, from consumers to customers to our Guests. They do a lot of focus groups, surveys, and then just general market research on that more extensive consumer base. But while the Guests are in the park, we are gathering their thoughts on everything they’ve experienced, and we’re also looking at their behavior during their stay such as where they stayed, the tickets they bought, and the dining reservations they had. And then we can also make educated guesses about how their vacation went based on things like: Do they rebook? Was that their 1st time trip? Do they come back for more? How does their following vacation change from the vacation that they’re currently having?”
Glynn Smith: “It seems like Disney is collecting data from a bunch of different areas on one specific customer to make their experience unique. Is that data transferable so a unique customer’s data may be used by another departments’ data science team to influence their experience on a cruise or a different resort location?”
Morgan Ferguson: “There are several barriers with sharing that data. For example, the streaming data is not shared with us, and vice versa. So that’s not something that’s currently happening on a broad scale. So it’s not like we see that you’re watching the Mandalorian on Disney Plus, and we send you a notification on the app to say ‘hey, we recommend Galaxy’s Edge’, right? It’s not to that level, but within the ecosystem of Walt Disney World, we can use some of that connected data to help shape the Guest experience. The most common way to experience Walt Disney World is to go through the app now. Most of our Guests have an account on the app. The Guest’s whole trip gets planned through that account, so they can link their Resort reservation, tickets, dining reservations, and more in one spot. If the Guest chooses to buy the Lightning Lane passes, then that would all be connected through their account on the app as well. So, of course, that data helps us determine the Guest behavior on their trip. So, as I mentioned before, we can look at the Guest experience holistically. So that’s how we would do that: by people connecting their experiences to their account on the app.
For example, we can say that Glynn went to Magic Kingdom on a given day. We know this because his ticket was for Magic Kingdom. He’s staying at the Grand Floridian and had dinner at the Be Our Guest restaurant – that kind of thing. We can see the Lightning Lanes you bought and try to piece together your behavior. But again, it’s all voluntary for people to link their data. So, it’s not the complete picture of everyone who walks through the gates, but it is a very large subsection of people willing to share their data with Disney. And it makes their experience so seamless that people are excited to have their whole trip linked like that.”
Glynn Smith: “What methods or tools do you use to work on the customer data? One thing we’re learning about right now is Bayesian statistics and its popularity in marketing and business. I was curious if you use it in your day-to-day.”
Morgan Ferguson: “Yes, a lot. I can’t go into too much detail, but people are interesting. It’s really fun looking at the data and finding outliers and trying to put yourself in that mindset and think of questions like how? Why did this person go about it that like? What’s the thought process behind it? And of course, that’s not where you focus your analysis. But when you’re sitting there in the office looking through data, it can be a fun little personal game to try to understand your Guests better.
In terms of the actual analysis and the projects that we’ve worked on, there are always a lot of surprising findings. And that’s why we do them – because a lot of times we’ll be in these big meetings with executives, and we will have to show how the data doesn’t support their initial thoughts. So, we end up doing a lot of interesting hypothesis testing. A lot of the driver analysis that we do is really fascinating to me, understanding the factors that lead to certain outcomes.
One example that I can tell you about is with after COVID, Disney World implemented what they call the Disney Park Pass reservation system. So, you had to make a reservation for the theme park that you wanted to visit on a certain day. That system is mostly phased out for our regular paid Guests now. It’s mostly just for the employees and Annual Passholders now, but with the Annual Passholders we were seeing unexpected ‘no show’ rates for these reservations. We would have X amount of allocations for Annual Passholders, and we found that some people just wouldn’t show up. They wouldn’t cancel their reservation or modify it, they just wouldn’t show up to the park. So, one of the questions that we had from our Yield team was how do we allocate the reservations to optimize the balance between the commercial side and the customer experience side? So, maybe we could overbook the number of allocations for Annual Passholders. But we don’t want to overbook too much, because then there are too many people in the park which lessens the customer experience for everyone. The wait times get too long, the dining reservations are taken, things like that. But then on the flip side, when we have so many people not show up, we’re losing out on some commercial opportunities. That’s people that could be in the theme parks shopping, buying food, buying drinks, etc. So that was a business question that came to us, so we initially did a driver analysis for it. We found that some of the factors that were leading Guests to not show up for their reservation were really fascinating, and not necessarily like what our leaders thought was happening. Then, based on the findings of our driver analysis, we were able to build a prediction model that forecasted the no show rate on a daily basis. This model predicted how many people, and which people specifically, may not show up for their Park Pass reservation. In turn, the team used these predictions to accurately over allocate on the number of reservations that they opened. So that’s been a really successful model that is still running today to make decisions on a daily basis of how many reservations they open.
Glynn Smith: “In a company where creativity and data often intersect. How do you see the role of storytelling evolving in your work as a data scientist, and what impact does it have on how data is used to enhance guest experiences and drive some decisions?”
Morgan Ferguson: “I would argue that the storytelling piece is the most important aspect of it. You can build the most incredible model, and it can have awesome results. But if you don’t present it correctly, and you don’t get buy-in from the people who are actually going to make decisions off of it and implement it, it goes nowhere and you get nothing out of it. That’s to me the biggest part of it – creativity is my favorite part of the job. You have all this data, but you have a lot of creative freedom in how you pull it and what data you use. Of course, there are legal implications that you have to consider. There are quality assurance factors as well. You can’t just pull some random data and not know how it’s collected or if it’s accurate. There’s a lot that goes into the process where you have to become a subject matter expert on the data that you’re working with.
You have to use a lot of creativity combined with business knowledge to go in and choose the right data and select the right model in order to answer the question at hand. In the end, the presentation part of your work is totally an art form in itself. It changes with every audience. For instance, if you’re presenting to the VP or your manager, you might word things a bit differently. It is crucial to know your audience really well and tailor the presentation towards them and their specific knowledge and needs.”
Glynn Smith: “What advice would you give to aspiring data scientists who are interested in working in a creative, guest-focused environment like Disney?”
Morgan Ferguson: “Continue to grow and develop your data storytelling skills and then really get a deep understanding of the business side of what you’re working on as well. Because that’s really going to help you decide what data you pull, how you can use it, and where to find it, too. You might have a massive data repository with all kinds of random tables, but if you don’t know how each one is being collected, then you can’t make informed decisions on how to use it. It takes a deeper passion in the subject matter that you’re working on, I think, to really elevate to being a great data scientist and to being successful at a company like Disney. Really knowing what you’re working on and how it affects the bottom line and the customer, such as what the outcomes and the impact of your work are, I think is the most important to being successful in a role like that.”
Columnist: Glynn Smith