Discovering the Sentiment Behind Loyalty Points Transactions in 2018
Author: Patrick Coddington BBA, MA, MSc
Patrick has been extracting actionable insights to drive commercial performance for over 17 years, for companies ranging from Expedia, to Dunnhumby, to the Government of Ontario. He continues to drive growth through innovative predictive analysis as Director of Marketing Analytics and Insights at Points International (TSX:PTS) (NASDAQ:PCOM), the global leader in loyalty currency management.
Probably the most widely touted benefit of Big Data is that it allows us to personalize customer experience to deliver the right offer to the right customer at the right time in the right place. This means using Big Data to build predictive behavioural models that anticipate what a customer may do next, based on their previous interactions and transactions.
In 2018, brands will look to better understand the sentiment behind loyalty points transactions for actionable insight into consumers’ sentiments towards their loyalty program.
However, as Martin Windstorm pointed out for Fast Company, algorithms are great at working out what customers are doing, but not at explaining the ‘why’ behind them – the sentiment behind the act. Therefore, one has to analyze the sentimental ‘small data’ to really understand the ‘why’, the feeling that drives a transaction.
By understanding the ‘why’ not just the ‘what’, it allows us not only to make better predictions but to discover hidden value and to tweak or even fundamentally change our offering.
Big Data gets personal
Recently, we have seen some novel approaches to understanding the sentiment behind customers’ transactions. TravelDataDaily, for example, collated millions of social media posts with Cathay Pacific’s member data. Using this, they were able to determine with greater than 75% accuracy members’ loyalty tier status and predict the likelihood of them switching programs.
As the global leader of loyalty currency management, Points International is in a unique position to see how members behave; we sit in the centre of the loyalty industry, watching members interact with different programs.
Around 20% of the transactions we handle are points/miles transfers: one member transferring his or her points or miles to another. We wanted to know what was making people so willing to give away their hard-earned loyalty currency. What were they thinking?
A goldmine of sentiments
The key was this: when someone transfers points or miles using our system, they have the option to add a personal message for the person they are transferring to. Just over 1/3 of people choose to include a personal message with their transfer, numbering in the tens of thousands of personal sentiments.
By parsing these messages and aggregating them with other member data, we were able to get some fantastic insights into the feelings behind these transactions, at the moment of transfer.
In their own words
A large proportion of these transfers appeared to be among family members, indicated by members sharing the same surname or address. The messages, perhaps unsurprisingly, indicated that these ‘Transfer Gifts’ were centred around life events such as birthdays, funerals, holidays, weddings, etc.
“Miles for mom’s trip to the wedding”
“Here are some miles from Dad to come home for Grandma’s funeral”
“Happy Birthday! Please put these miles towards something wonderful!”
There were far fewer transactions associated with financial or account-based sentiments (such as miles expiring or changing programs) than we had expected, although there were a small percentage who indicated expiring miles as the reason for the transfer. This goes against our own research from a few years back, which indicated these were the key reasons for transfer. These messages also held rich ‘small data’ sentiments that spoke volumes: “Transferring miles because I’m stationed in Japan for 2 years and cannot use before they expire”.
Big insights from Big Data
Previously, we were able to answer the who, the what, and the when. With this rich data source we’re now able to answer the all-important why as well.
By analyzing this kind of ‘small data’ with a Big Data approach, we unearthed a wealth of member sentiment using this unstructured data, from a previously impractical source, as it was unfeasible to manually read, sort and analyse each of these 30,000-or-so messages.
This Small Data approach has allowed us to come up with even better predictors of when members are likely to transfer (e.g. upcoming miles expiration, or a birthday). More importantly, we’ve been able to supercharge our marketing messaging to be even more relevant to our members, changing mile transfer transactions from functional to emotive. “Transfer now!“ marketing messaging becomes “Miles for mom’s trip to the wedding!”.
This kind of natural messaging will become increasingly important in 2018 and beyond, as voice-activated search and devices become more commonplace.