AI Analyzes Billions of Data Points to Tell Cruise Line What Matters Most
Data has become one of the biggest commodities for businesses, but data without analysis, context and action doesn’t offer returns. Celebrity Cruises, part of Royal Caribbean Cruises, is hyper-focused on data to not only market to its existing and potential customers, but to make sure that marketing dollars are not siphoned off by fraudulent activity.
HT sat down with Matt Maule, VP of Business Intelligence, Celebrity Cruises, to find out more about Celebrity’s approach to data and how it is part of a strategic plan to gain competitive advantage.
There are many aspects of BI and analytics when it comes to marketing, but a major part of Celebrity’s CRM strategy is one-to-one marketing. Maule notes that this requires the loyalty team having the ability to roll data up to the CRM to have a complete picture of cruisers and potential cruisers.
Celebrity Cruises leverages an automated business analysis (ABA) platform from Outlier, which quickly sifts through volumes of data to identify changes or business shifts that are most important.
“We have a lot of first party data that we collect,” Maule says. “The transactional data allows us to manage bookings properly. We have enough information so that when a guest rebooks, we don’t have to repeatedly ask for the same information. We want to be able to recall information when it’s appropriate.”
This data also includes deep transactional data that occurs prior to a cruise, such as shore excursions, spa bookings, pre-ordered beverage packages. Maule explains that = data is captured and combined with information on what, when and how they pay to formulate a fuller picture of guests.
“We will then segment our guest profiles based on that,” Maule explains.
Appraising guests’ commercial interactions allows Celebrity to market to them appropriately, being able to gather and leverage analysis around buying habits. Such analysis pulls in billions of data points that need to be culled down to actionable, meaningful insights, a task best suited for the capabilities of artificial intelligence.
This is where the Outlier partnership comes in to play. Maule says he was seeking a tool that would be able to help provide a clear picture of the thousands of data points Celebrity has to look at including: 30,000 travel partners in the U.S., 600 to 700 open sailings for sale, 30+ stateroom categories and an immense number of different price programs and promos. He wanted a tool that not only showed the team all of the things they could be looking at, but also what they should be looking at.
“Analytics is a mindset,” Maule says. “If you’re not a decent analyst, there are so many different conformational biases and analytical techniques that can be applied to interpreting data. With the democratization of data and the proclivity of dashboards and reports, that it is important to focus on the right things. There are things that matter, that could be moving almost imperceptibly.”
The Outlier tool identifies what normal behavior looks like, and when something happens outside of normal, it will report if the actual performance is over- or under-performing outside of the expected range. It does so in a simple, graphic visualization that makes it easy for team members to interpret the data.
Catching & Reacting to Ad Fraud
Digital ad fraud has become a big issue as marketing has shifted to online channels. Celebrity is able to track when advertising has hit in certain regions because web visits will spike in specific regions. Outlier also offers more detailed web traffic data when there is a strange referring domain that pops. This is when a rogue domain will act in an unethical or illegal way and send an email to millions of people to go to a URL to get a fee on a cost-per-click basis.
“Thanks to Outlier, we’ll be alerted to that abnormal activity in less than 24 hours and be able to report it to the proper authorities to have it shut down,” Maule explains. “This way we don’t waste money on fake visits.”
On the positive side, Celebrity is also able to see the impact of short-term incentives and determine the type of uplift promos through trade partners have yielded across various channels.
“We’ll see a rise in demand and adjust accordingly,” Maule describes. “Most importantly, we might see an uptick in results but won’t over-celebrate success on one hand, or panic if performance dips as long as it is in a typical range.”
Trend detection is a valuable feature for Maule and a differentiator from other business intelligence tools. When the system detects a change away from normal behavior, Maule’s team can drill down to key drivers while accounting for factors such as seasonality. The AI then automatically draws associations or identifies relationships with other data points that may be performing in the same way.