Loyalty programs are data collection machines. Across the globe, many companies are already experimenting or heavily using either machine learning or artificial intelligence models to support operations and gain invaluable insights and efficiencies.
For companies wanting to dip their toe into this space, the blog article details some use cases and examples for consideration.
Use case 1: Marketing Communications
Machine learning/AI-powered marketing execution can automatically construct and send individualised communications to members across a variety of channels, then track responses and evolve future communications to optimise engagement potential. Their ability to serve up individualised communications to members across a variety of channels makes them highly effective, while helping to reduce marketing costs.
AI-powered marketing execution is primarily the domain of larger loyalty programs due to the cost and the enormous amount of data required to optimise the algorithms.
A May 2019 Credit Suisse report indicated that major supermarkets:
‘Are almost uniquely positioned to leverage their growing digital capabilities to establish a competitive advantage in the promotional expenditure component of the food retail value chain. With material efficiency benefits feasible, it is likely that growing digitally led efficiency will support market share gain and outperformance’.
This includes the ability to offer discounts on products to members who are price sensitive, while charging full price to customers who regularly buy the product, with machine learning/AI tracking the campaign’s success and adjusting accordingly.
Example: Woolworths Everyday Rewards
Woolworths utilise Quantium’s AI platform to deliver individualised communications to their 10m members. This includes mixing items the member hasn’t bought before with past purchases to increase familiarity and propensity to purchase.
Use case 2: Personalised Digital Experiences
Another area where AI platforms are being applied to hyper-personalise member experiences and communications is websites and apps, which is also a big source of member data.
AI platforms can identify individual members when they are engaging with a range of digital channels, access their profile information (encompassing hundreds of datasets) and automatically trigger a relevant promotion to stimulate a desired behaviour from the member.
One advancement is AI which reads human facial expressions online to determine nuanced emotions and complex cognitive states. This allows digital technology to recognise member sentiment and respond with an appropriate digital experience.
Global player Tealium provide AI capabilities to Domino’s to grow online and app order spend, as well as the completion of transactions. The capabilities have enabled the business to empower stakeholders on an international scale to make more timely data-driven decisions for user segmentation and marketing campaigns.
Use case 3: In-Store Personalisation
Numerous retail stores have begun to experiment with AI in ways which have the potential to change the traditional retail experience for good. AI is being applied directly to the in-store experience, whereby members can receive real-time recommendations, virtual assistants, and test products without physically trying them on.
This level of hyper-personalisation in-store enriches the member experience and aids purchase decision stimulation.
Example: Kroger
Supermarket chain Kroger partnered with Microsoft to create a personalised shopping experience with smart shelves. With the Kroger app open on their phone, members can walk down the aisle and sensors will highlight products the member might be interested in buying based on account information and their previous shopping history. It also provides dynamic personalised pricing and will show ads and point out items of interest, making the shopping experience efficient and convenient.
Example: Walgreens
Walgreens are using AI to tailor ads for different types of shoppers on in-store cooler doors. The “smart” displays have cameras and sensors that draw in data, the AI analyses it and then relevant content is displayed on the screen. The doors can also detect which items are looked at or picked up, creating a feedback loop for the AI to learn and personalise even further.
Use case 4: Virtual Try-Ons
Consumer brands are rolling out virtual try-on technology to allows members to try on clothes, shoes, cosmetics and more using the webcam on phones or laptops, or screens installed in stores.
The technology allows brands to load thousands (or even millions) of items into the platform for members to try. More advanced platforms can utilise member data to generate general recommendations, as well as suggesting items which complete their wider wardrobe.
Example: Mac Cosmetics
Mac Cosmetics utilise this technology, where members can try on 200+ shades at home. The positioning states that meeting your next favourite eye or lip colour online has never been easier. Members can instantly swipe on 200+ shades with our new virtual try-on, using three easy steps.
1. Choose a product and click “TRY IT ON” on the product page.
2. Enable your live camera, upload a photo from your device or choose a model.
3. Scroll and click on swatches to instantly see any shade on you.
Use case 5: Conversational AI
Conversational AI refers to the use of messaging and speech-based virtual assistants to automate communications and create personalised customer experiences at scale. These technologies are fuelled by the rise of messaging apps, voice assistant platforms and general advancements in AI, including deep learning and natural language understanding. The beginning of this technology was really your basic chatbot.
True conversational AI is the next step where machines will be capable of understanding and responding to human language and behaviour in real-time, learn from these experiences, and then apply these learnings to future conversations. These intelligent assistants will not be limited by the functions anticipated by a programmer. This technology will present new ways for brands to appear more lifelike, respond contextually and personally at scale, and shift the brand-customer dynamic from many-to-one to one-to-one.
Example: The Cosmopolitan of Las Vegas and Rose The Hotel Chatbot
The Cosmopolitan of Las Vegas is a luxury casino and resort which encourages hotel guests to interact with their virtual assistant during their stay. The hotel chatbot, known as Rose, is designed with a unique voice with over 1,000 conversation threads to offer guests ways to book experiences such as restaurant reservations, spa treatments, events tickets, and other exciting adventures.
Rose can provide recommendations based on what the guest desires and provide insider information like secret menu items to drive guests to specific bars, clubs, and restaurants. For example, a guest can ask ‘What should I have to drink?’ and Rose may make a special drink recommendation which is not on the menu, both assisting and delighting the guest.
Use case 6: Security & Fraud Monitoring
In addition to enhancing the member experience, machine learning/AI are being applied to protect programs and their members from security breaches and fraud.
The advantage of an algorithm-based monitoring approach is speed, volume of account reviews, and the ability to learn and adapt to changing circumstances.
Example: GetPlus Indonesia
GetPlus is a major points program, where members can scan their receipts from participating merchants to earn GetPlus reward points through the app via OCR. GetPlus set up configurable, action-based scoring rules to evaluate the risk of loyalty fraud in real time. If a high-risk earn order is identified, it is suspended by the AI to be reviewed by human operators.
The fraud-detection capability also provides reports to be actively monitored, and orders can be analysed by risk status. It can also make modifications to scoring algorithms as patterns of fraud change.
AI Marketplaces
As a consultancy, we are talking to more and more AI platforms to understand just how robust some of the new technologies are. We have explored dozens, with some super interesting use cases that can be applied in really cool ways.
Veritone, for instance, are effectively an AI marketplace that aggregate lots of different start-ups in the space and basically sell their services through them. In fact, they currently have about 200 companies selling through them. They also sell their own native models.
So, a company can simply approach Veritone, give them a use case they’d like to solve and Veritone will utilise its network to then present a solution and if suitable, execute on it.
Summary
AI and Machine Learning can be applied across and used to support marketing communications, personalised digital experiences, in-store personalisation, virtual try-ons, conversational and security & fraud monitoring use cases. Any company that runs a loyalty program will, at some point, have needs across most of these areas.
AI and Machine Learning is also significantly more accessible than before, with many platforms and start-ups playing in the space. These same platforms are quite open to running proof of concepts and pilots to prove that the technology works and gain a competitive advantage.
If you’re a loyalty program marketer and looking to allocate some budget, this is definitely a worthwhile space to explore.
