When you’re in the hotel industry, even the absolute most loyal of loyalty members will stray to the competition from time to time. Figuring out how to identify and combat this tendency is increasingly falling to technology, which has made dramatic leaps and bounds in data discovery when it comes to increasing loyalty member walletshare.
One technology that is driving progress for hotels in this arena is the Big Data powered “Customer ConnectId™”, a solution built by InfoTrellis to gather and match online social profiles to loyalty member profiles.
Like any new technology, the proof is in the pudding – and that pudding is the first customer use case. With this particular identity resolution technology, the first use case is taking place within the prestigious offices of one of North America’s leading hotel chains.
In the hands of the hotel’s analytics team, the ability to connect loyalty member profiles up with their public online activity (like Twitter accounts, Facebook accounts and travel website reviews) became a powerful tool in their search for better strategies for increasing customer loyalty and, ultimately, loyalty member walletshare.
Initial work with the massive new sources of data, automatically matched and consolidated into customer records, provided fascinating information about how people indicate travel when they interact with social media. The hotel’s analytics team discovered:
- Comparing customer check-in rates to those same customers’ FourSquare and Yelp check-ins with competing hotels, they were ahead of their competition
- BUT, comparing customer check-in rates to those customers’ FourSquare and Yelp check-ins with their own hotels, they only broadcasted about 50% of actual check-ins, suggesting that:
- The average socially active hotel guest only checks in to FourSquare or Yelp about half of the time
This meant that while at first glance they seemed to be retaining their loyalty members with a high degree of success, more rigorous scrutiny revealed that they were likely staying with the competition as much as 50 – 60% of the time when they travelled.
To provide a higher rate of accuracy when it came to pinpointing when their loyalty members were staying with the competition, the analytics team expanded their search to include check-ins at airports and restaurants far from the person’s hometown – both good indicators of travel. They also started leveraging the technology’s Natural Language Processing capability to incorporate informal, unstructured “check-in” messages, like a casual mention or reference to travel in a tweet or an update.
The end result of all this hard work is that now they have a much clearer idea of which of their loyalty members are staying with the competition and how often and, eventually, will be able to build models that help them to understand why they chose to stay with the completion. The immediate benefit is that their loyalty and marketing teams now have an unambiguous target list of loyalty members who travel frequently without staying with them – an obvious opportunity for the most growth with the least effort, if they can target them with the right offers and messages to win back a greater share of their business.
The next steps for their test of the technology will be to build out the depth of the profiles by pulling extra dimensions from the social data that can be used to understand their motives, behaviors, and preferences.
By employing a machine learning algorithm they expect to be able to get as much information as they can about their 10+ million loyalty members and then, using this complex pattern identification technology, accurately predict things like hobbies, interests, life events, job title and family status for those loyalty members who don’t list them explicitly. From there, ten million loyalty members suddenly become much easier to understand and custom-tailor offerings and experiences for.