Improving customer focus with intelligent systems
Industry groups everywhere have a lot to gain by catering to today’s phygital consumer, a savvy shopper who combines online and in-store approaches to planning, purchasing, and packing for a travel adventure. To align their businesses and their IT systems with the modern traveler, many companies are engaged in digital consolidation programs. These programs have the potential to keep travel relevant and rewarding despite unforeseen changes in travel schedules and the priorities of individual travelers.
Let’s look into the future to see where successful digital consolidation programs could make the following scenarios commonplace:
- Jump on discount travel offers with a HUD. A customer shares with an outdoor sporting goods retailer that he likes to hike and considers Greece a favorite destination. One morning as the customer makes his daily commute to work, the transparent head-up display (HUD) that overlays his windshield lights up with a holiday package offer for Greece from the outdoor retailer. The offer includes discounted travel and accommodations, a three-day schedule of possible hikes, and an equipment list. An overlay shows the store location where he can pick up the equipment on his way home from work. The HUD lights up to display the customer’s verbal confirmation and adds a reminder when he’s on his way home.
- Travel offer overcomes budgetary restrictions. As a savvy phygital shopper, our customer with the HUD system has registered his monthly financial plan with his banking system. He chooses a list of budget options to appear on his screen when he receives an offer. Therefore, the offer for a holiday package trip to Greece comes with suggestions: purchases he can move to next month and orders he can cancel to accommodate the trip. He chooses to postpone his golf club order and proceeds with the purchase of the trip.
- Keep travel plans intact despite in-flight schedule changes. A flight from New York to Frankfurt will experience a delay on the ground in Frankfurt due to a road traffic incident that prevents the caterers from getting to the airport on time. Aboard the flight is a group intending to travel onward from Frankfurt to a one-day stopover in Dubai before proceeding to Mumbai. The airline’s systems predict the impact of the traffic jam and send an alert to the controller’s desk with possible rebooking options. By the time the plane lands in Frankfurt, the airline has revised the group’s travel plans, booked alternative accommodations, and sent revised boarding permits. The flight crew receives an in-flight alert on their tablets with updated schedule instructions to pass on to the group. The group continues to their original destination without missing out on the holiday stop in sunny Dubai.
Apart from obvious integration challenges, the above scenarios contain some critical differences from what constitutes business-as-usual today.
- Collaboration between retailers, industry verticals, and providers. Instead of going to an airline’s site to book a flight and to a hotel’s site to book accommodations, the customer in the first two examples booked both through a retailer of outdoor products.
- Shift in focus. Rather than just trying to make their products as appealing as possible, the outdoor products retailer focused on understanding and prioritizing customer desires while fitting their products into a pre-integrated bundle.
- Change from request-response to predictive approach. In a traditional request-response system, a user initiates an action, a request goes to the server, and the system replies to the query with a response. In our scenarios, the application initiates the offer, not the other way around. The applications are aware: they knew who the user was and probably knew the likelihood of the user accepting the offer as well.
- A different level of awareness. To increase the relevance of an offer and the probability that it will be accepted, applications need to apply various aspects of machine learning. Machine learning is a branch of computing that explores algorithms capable of learning from data and making predictions based on that learning. This differs from traditional rules engine implementations, which take a request-response approach based on static program instructions.
- Predictive analytics can identify the impact of the traffic jam for the group in the airline disruption example.
- Supervised learning and unsupervised learning with various sample datasets prior to application rollout can provide the learning steps necessary to gauge whether our would-be traveler would be likely to buy a trip to Greece. The application uses these steps to gauge customer likes/dislikes and context (upcoming and recent events in his personal life) that might impact his travel decisions.
- Two forms of analytics – machine learning and data mining – need a big data store, or “lake”, to act as a repository for the details involved in determining an offer. Machine learning focuses on prediction and is based on known properties learned from training data. Data mining focuses on the discovery of previously unknown properties in data.
You’ve just had a taste of how selling opportunities will change in the near future through intelligent systems. In the era ahead, applications and systems will be more intelligent, in some ways, than the programmers who created them. And both the customer and the business will benefit. We’d like to hear your feedback on this.