Winning Brand Preference in Low Loyalty Segments: Contextual Personalization

I recently read an interview with Jennifer Wilson ( Lowe's Companies, Inc. CMO) and the Retail Brew team about the challenges of building brand preference in the home improvement sector, for which purchase decisions are driven primarily by convenience instead of customer loyalty. According to Jennifer a staggering 65% of consumers aren't loyal to any one home improvement retailer. They’re happy to shop around, picking up supplies from whichever store is most convenient or offers the best deal at the moment. If you work in retail this might sound familiar to you.?

This is a challenge that exists across the landscape of all sub-sectors within the omni-channel retail space (And many outside of the retail space as well). Consumers are bombarded with choices and can easily compare prices and products across multiple channels, this coupled with the level of access that e-commerce provides has led to a rapid decline in consumer loyalty.? According to 麦肯锡 research around 75% of consumers have tried new shopping behaviors since the pandemic, with 39% switching brands or retailers, and most intend to continue exploring these options.

While this might paint a challenging picture for retail, it’s not all doom and gloom! With the level of access consumers have today, achieving brand loyalty is more challenging, but not impossible. Customers want to feel valued and understood, and they're more likely to engage with brands that can provide them with tailored experiences that meet their individual needs - so much so that, 埃森哲 research shows that 91% of consumers are more likely to shop with brands that provide them with tailored messages and offers.? Before you start rushing to add your customers' first names to every email subject line, hold up! While personalization is key, the old-school tactics of sprinkling in merge tags like magic dust won’t cut it anymore. In fact, it might even backfire. Nobody’s impressed by seeing their name awkwardly plastered across an otherwise generic message (we’ve all seen those cringe-worthy emails).

Traditional personalization in lifecycle messaging, like using a customer’s first name in an email subject line, is totally broken. Research from 麦肯锡 , Yieldify , Bluetext , and others shows that traditional rules-based personalization has lost its impact and even could backfire, as it relates to brand loyalty. So if traditional personalization methods are falling flat, where do we go from here? The answer lies in marrying customer profile data and generative AI to create “contextually relevant” personalization. Instead of simply pulling in names or favorite products, let’s use those things to inform our language and messaging.

So how do we do that? Well let’s start with the data collection. Jeena Sharma aptly points out that loyalty programs are a great way to create a fair trade of information for monetary benefit. This is likely the reason that loyalty programs are on the rise - this year (2024) 63% of executives reported an increase in their loyalty program budgets. Now that we have our data, let’s talk about operationalizing that data!

The secret to operationalizing this data, at scale, is to bring enterprise-grade generative AI into the equation. Jacquard brings this approach to life with its Contextual1 Engine. Their engine leverages native language contextualisation using natural language generation, that is? optimised using a Contextual Multi-Armed Bandit (CMAB) model. This model doesn’t stop at basic personalization, instead it factors in additional context—such as user characteristics, time of day, and past interactions—to decide how to generate the language and which language variant to serve each user.?

Here’s how it works in practice: if a user often engages with content around specific themes, the system generates natural language based on that user data and product data. From there the CMAB dynamically adjusts to serve the language variations that resonate most with each user, boosting the likelihood of engagement and - eventually - brand affinity.

The key advantage of the CMAB algorithm is its ability to draw from real-time data, leveraging each user’s engagement history to tailor language and messaging on the spot. This approach goes beyond broad trends; refining messaging to fit each user’s unique preferences and behaviors, creating a highly personalized experience. By considering both overall performance and specific engagement patterns, Jacquard’s CMAB approach allows brands to meet customers exactly where they are, offering a level of personalization that truly resonates.

In today’s competitive landscape, this type of contextual, data-driven personalization is the way forward and is crucial for brands aiming to establish meaningful customer loyalty - especially in low loyalty segments.