Airports are an interconnected system where one unforeseen event can tip the scale into chaos. For a smaller airport in Canada, data has grown to be its North Star in an industry full of surprises. In order for data to bring true value to operations–and ultimately customer experiences–those data insights must be grounded in trust. Ryan Garnett, Senior Manager Business Solutions of Halifax International Airport Authority, joined The AI Forecast to share how the airport revamped its approach to data, creating a predictions engine that drives operational efficiency and improved customer experience.
Here are some highlights from Paul and Ryan’s conversation.
Building a data culture
Paul: You joined Halifax International Airport Authority over a year ago. Tell me about what you were trying to build or replace or accomplish.
Ryan: First, I wanted to build a culture. Data needs to be an asset and not a commodity. And we need to think of it differently as something that we leverage and value. But the reality is people weren’t valuing it. I’m reminded of a previous place where I worked in finance and reported to the CFO. It was early on in my time there and I was getting to know him. He tells me, “Ryan, I don’t value data.” That obviously stunned me. I just joined this organization to do data, and you don’t value it. He elaborated and explained the reason why is because he could ask five different people the same question and get back five different answers. That always stuck with me.
Building that culture is around trust. Why am I doing this? Why are we doing this? What’s the reason for data? Everyone may answer and say, “informed decision making, generate profit, improve customer relations optimization.” That’s not it. You’re in the business of building trust. That’s your business. As soon as you build trust, you can do anything.
Paul: Where do you see your data journey heading? How did executives feel about the data refresh?
Ryan: Kudos to our executives because they bought into the value of data from the start. We didn’t need to sell them on why we needed it. They gave us the opportunity to build it. We have a strategic plan like most organizations, but the underpinning part for me is to change the culture. I want to get people to think of not what has happened but what could happen. If it did happen, is it significant? Just because it seems like a big number to you, it doesn’t mean it’s actually significant. Take that gut feeling and don’t throw it away, but just pause for a second.
For example, we send routine reports to the senior leadership team. After one particular report, our CEO asked why a particular number was down. We said, “Do you remember that day? That was the hurricane.” You could see that relationship. People began to get that understanding that external forces are pushing numbers. Was that dip in numbers significant? Not all the time, but that’s why we support this broader thinking with data so people can plan for erroneous events and better understand the shifts.
Transforming operations and customer experience
Paul: Talk us through how you’re using that data—like passenger transit data—to plan for future events.
Ryan: Instead of looking in the past, we’ve built a predictive model and its origins come from people trusting in us—they ask us about different scenarios. Victor, who’s in airport planning, asked us to help his team understand when someone might show up. To dive into the problem, we had to uncover what that means for him. He specifically asked to know in the future what our biggest day is going to be or our biggest hour. From there, we have data we can look at, but we also have the schedule: When does a flight come? How many are we expecting? What size planes are they? How many seats do they have? And you can’t just say, this type of plane has these many seats. It depends on the carrier. They might take them out to give those nice comfy seats.
We also pulled in real-time weather bulletins. And then we partnered with Halifax Discovery. That’s part of the group that brings in events into Halifax. That partnership gave us data of when people are coming 10 years in the future. This ultimately can build a model to look at the history to determine, “Hey, every Wednesday it seems to be this big and there’s these many planes.”
But what about when the snowstorm hits? What did that look like? What is the impact over a long weekend? We leveraged a model put out by Meta and it predicted it out and it predicted it pretty decently. There’s so much more we can use with this model. We can even use it to figure out how to staff security because we have a pretty good idea when it’s going to be busy.
Paul: What’s the next big innovation for Halifax airport?
Ryan: I think the big one for us and for me is building on passenger experience. Travel anxiety is a real thing. People get extremely anxious about traveling. My hope, my desire, my dream is to take a data-driven approach and provide something to the general public that helps them lower that anxiety as low as possible when their journey starts by opening their door.
Don’t forget to tune in to Spotify or Apple Podcasts to listen to future episodes of The AI Forecast: Data and AI in the Cloud Era.
