Keith Dawson's Analyst Perspectives

There is an industry assumption, call it a consensus, that increasing the degree of personalization in customer interactions is beneficial: it improves loyalty, lifetime value, CSAT and other key metrics. Core to this assumption is that it benefits both the customer and the business. Whether or not that’s actually true is somewhat beside the point. The fact is that the industry has identified a use case that seems to appeal to buyers, so the next step down that road is for software providers to figure out the best way to articulate the capabilities and the rationale for adoption. How can this be communicated to potential buyers, and how can they be helped to see hyper-personalization as a successful strategy for them? Personalization is one of those things that straddles the fine line between useful and creepy. That’s from the end-user consumer point of view; from within companies running customer experience (CX) programs, it can look instead like a solid way to boost loyalty and, one hopes, revenue. That means that the case presented to buyers should rest on balancing those views. So, software providers have to provide buyers with arguments for why it works inside companies while also benefitting customers sitting outside. This is especially important as enterprises look closely at how to boost loyalty—by 2029, one-half of enterprises will focus on customer value and loyalty as the key metrics defining success in customer relationships. Consumers care about things like privacy and control. That suggests building in levers they can use to dial up or down the personalization, along with enough obvious (not hidden) opt-in/out mechanisms. Whether or not they take advantage of those controls, merely including them may encourage consumers to see personalization as a benefit rather than an intrusion. For brands and enterprises, it makes sense to root your messaging in descriptions of value: how it enables the development of a level of trust and loyalty from customers that can be leveraged into positive actions like customer advocacy or longer lifetime value. In a 2022 ISG survey, 77% of enterprise respondents said that personalizing services based on individual buyer context was extremely or very important to their organizations. And 79% said the same about providing real-time recommendations based on purchase history. (That’s another proxy for personalization.) In the years since that survey, I’m confident that advances in artificial intelligence (AI) and other tech have only made delivering personalization even more attractive to businesses. And yet in 2024, only 26% of enterprise respondents said they were already invested in AI projects related to personalization. That suggests a gap between a desire to use data for more personalized activity and having tools focused directly on that use case. If the technology is ready, and the consumer base can be made comfortable, this may be an opportunity to extend features in a new and interesting direction. Things look slightly different when you pivot from the general benefits of personalization to whether it produces concrete results related to loyalty. Brands and consumers have very different ideas about what loyalty means and how it is expressed. Brands care, of course, because loyalty is a monetizable proxy for long-term customer intentions. For consumers, though, loyalty is much less of a direct concern: it’s easier for them to find a new business to work with than it is for the business to find a new customer. Customer expectations related to loyalty are a mixed bag. They appear (anecdotally) to have less “loyalty” to brands and businesses than in the past, depending on how you define loyalty. (If you define it as “affinity” or desire to recommend/advocate on a brand’s behalf, you don’t see much of it. If you define it as an ongoing positive relationship with a business, in which the customer engages repeatedly without negative effects, then it’s clear that customers are pretty loyal, whether they consider themselves to be or not.) My view is that customer behavior is like lanes on a highway—they tend to stay on course unless something pushes them sideways. Behavior is more of maintaining a pattern of doing what works, and if you consider that loyalty, then it’s pretty high. So, the biggest challenges for brands are in definition and measurement. To foster a sense of affinity in customers, you have to a) provide them with what they want and b) make them aware of the fact that they want it and are getting it, without incident. Brands can do surveys to take the temperature of customer sentiment, but those are often unreliable indicators. Modern AI sentiment analysis is better but is really a tool for telling whether someone is satisfied in the moment more so than with the long-term relationship, which is what you’re looking for in a loyalty measurement. Other scoring tools, notably NPS, try to get at the more nuanced aspects of loyalty. Because the relationship between companies and customers is inherently lopsided, businesses should understand loyalty more as a light commitment to continue that connection than as a way to turn detractors into advocates. Personalization can certainly improve loyalty, when used carefully. Think of personalization as a series of concentric circles. At the center is the most basic: knowing the person’s name, using it and referring in communications to specific aspects of the customer’s transaction and usage history. Everyone expects this, it’s not at all controversial and it’s easy to automate. It probably has very little impact on loyalty or the repeating of business since people see it as a basic courtesy. (It can be overdone via excessive unnecessary communications, but that’s not the fault of the personalization, necessarily. It’s a fault of the way the campaign is structured.) The next circle beyond personalized communications would be personalized recommendations and content. From the consumer standpoint it can seem intrusive, and it calls to mind examples that everyone can cite where they search for something online and then for days they keep seeing ads for products following them across the web. The two ideas may not be literally connected, but they suggest to the consumer (perhaps on a subsurface level) that brands will do or say anything to get their attention. How do you make this less intrusive? By being careful about how you tune your recommendations and by using contextual data about recent customer histories in order to avoid, for example, trying to upsell someone during a series of difficult service interactions. Or tailoring recommendations to a customer’s social media feed. Further personalization circles would extend the idea to making service or support directly tied to things you know about the customer. For example, matching customers to service reps with specific skills related to their histories or providing customized follow-up to interactions that provide tips or instructions, reminders of the need for maintenance or offers tailored to their specifics. A lot of this happens behind the scenes, invisibly to the customer, so there isn’t a lot of danger here of being creepy or intrusive. Personalization ultimately plays a role by encouraging the continuity of the relationship—if we define loyalty as the ongoing nature of a positive relationship, then targeted efforts like those described can keep the customer from straying to a different brand, keeping the customer engaged enough to be satisfied but not too engaged that they actively have to think about whether they want to do business with this company. I’m not sure that there is one specific investment that can make a big difference in customer loyalty. Many tools now being developed will have some impact on it, especially AI communication tools and more advanced analytics. If you buy into the idea that not doing anything to derail a relationship matters more to loyalty than coddling customers (and I do), then the best thing a company could do would be to streamline communication processes. For example, it is important to use personalization to understand and then deliver on customer preferences for channel choice. Or to ensure that measures of sentiment and satisfaction from conversational self-service channels are correlated with those from live interactions and past transactions to build a picture of “loyalty” that reflects what the customer is likely to do in the future. It is important to move past satisfaction as a metric because it doesn’t tell you much about the customer’s future behavior, and therefore, the revenue impact of loyalty. Enterprises need to view personalization as a mechanism to delivery loyalty, which is not an end in itself, but rather a proxy for understanding future revenue. Customers are not demanding personalization, but smartly deployed, it can make them more likely to continue to do business. Regards, Keith Dawson

Keith Dawson's Analyst Perspectives

The Tricky Relationship Between Personalization and Customer Loyalty

Adobe Applies Agentic AI to Customer Journeys

NICE Orchestrates AI and Automation for Customer Journeys

Our Analyst Perspective Policy

Our Analysts

David Menninger

Jeff Orr

Keith Dawson

Mark Smith

Matt Aslett

Matthew Brown

Robert Kugel

Stephen Hurrell

ISG Software Research

Subscribe to Email Updates

Posts by Month

Posts by Topic

Analyst Perspectives Archive