Someone recognize me!: Generative AI and loyalty rewards programs
How GenAI can be used to transform loyalty rewards programs and support DTC businesses in their struggle for greater profitability
Things were busy, but they are calming down. And so, I’m back, finally, to write that part two I promised on loyalty programs and generative artificial intelligence (GenAI).
A few weeks ago, I wrote a piece on how Web3 has expanded capabilities within loyalty rewards programs by using NFT technology to generate ID codes and metadata that separate you from anyone else. In that piece, I talk about Blackbird Labs, which raised a $24M Series A last week! Â
This week, I’m going to dive deeper into why loyalty programs are especially important to consumer companies now and how GenAI technology can be used to transform rewards programs and, subsequently, business outcomes.
As I mentioned in my first piece, loyalty rewards programs are one of the most important revenue drivers for consumer businesses. But while the average American is a member of 16 loyalty programs, usage rates are below 50%. A strong loyalty program with high customer engagement increases transactions per member by 3.5x and revenues by ~20%. This means that if a business can create a differentiated loyalty program with high customer engagement, it will have a significant positive impact on that business’ revenues and overall profitability.
Loyalty programs are becoming even more important today as the cost to acquire new customers (CAC) heightens due to an increase in paid advertising by platforms like Meta, a declining effectiveness of those ads due to increased privacy controls, and also a pressure to develop an omnichannel presence early. Increasing costs are directly impacting the profitability of DTC businesses. While businesses like Warby Parker or Away were able to get onto your radar by buying cheap ads on Facebook and then Instagram, younger DTC businesses like, most recently, Parade are struggling to prove significant returns due to the large costs of the DTC business model.
As the cost to acquire customers increases, it is ever more important to keep those acquired around. A strong loyalty program increases the lifetime value (LTV) of each customer by driving increased customer activity (visits to site, frequency of purchase, size of orders) over a longer period.
But what makes a loyalty program good? The number one factor is personalization.
In order for a business to entice you, it has to provide something you actually want, when and where you actually want it. The competitive advantage between businesses comes from the ability to gather and then effectively utilize personalized customer data across the entire customer journey and at scale.
A really strong loyalty program has to do the following for each customer: 1) understand her habits, 2) accurately identify her desires, and 3) generate enticing offers based on those desires.
GenAI, with its ability to use large language models (LLMs) and natural language processing to rapidly analyze data sets and then generate unique content outputs, unlocks new capabilities across all three of these areas.
Let’s break it down.
1 ) Understand habits: Customers prefer omnichannel shopping experiences (i.e., experiences that span across both brick & mortar stores and e-commerce sites) and so habits must be gathered across all platforms and then traced back to the same customer. While email addresses (e.g., giving your email at checkout) are an easy way for businesses to consolidate data behind specific customers, data gaps (i.e., data that doesn’t integrate across systems, data that is unusable because of errors) still appear. However, LLMs can automate data transformation and maintenance to streamline the integration process and resolve errors, as done by companies like Amperity, Twilio, and Lume AI. AI can also be used to identify strange habits, alerting businesses of potential fraud.
2 ) Accurately identify desires: Once a customer’s habits are established, playing to those habits is an easy way to predict customer desires. LLMs can further train on customer responses to promotions to refine offers and ultimately improve hit rates.
3 ) Generate promotions: This is where the new capabilities provided through GenAI truly come in. GenAI technologies can use insights into a customer’s habits and desires to rapidly create tailored promotional materials. An example is offering a price-sensitive customer a discount on the shoes they were looking at before they close the tab. Examples of businesses in this space include adcreative.ai or text-to-image programs like Dall.E2. Google has also begun to enter the AI-driven, ad-campaign generation space.
Established companies are already using AI to propel their loyalty programs and are seeing results. The Starbucks loyalty program, Starbucks Rewards, offers personalized promotions based on past ordering habits. Sephora’s Beauty Insiders uses AI to analyze customer data to provide customized product recommendations. Sephora also recently launched a Visual Artist feature which uses AI to scan a customer’s face and provide customized product recommendations. However, I have not yet come across a start-up that leverages GenAI to build a one-stop-shop loyalty infrastructure for consumer businesses in a way that Blackbird does for restaurants using Web3 tech. Given the significant revenue uplift provided by strong loyalty programs and also the cost-saving benefits through reduced fraud and expanded capacity for the marketing team, I do feel there is demand for this product if created. And so, if you’re building this product that supports all three features of a loyalty program that I outlined above, I’d be interested to hear more!