Personalization has become an overused buzzword. Every company claims to be doing it. But are they? What exactly is personalization? Some might define it as, "Something built or delivered for me, just the way I want it." But companies have been doing this for a long time: offering the choice of fabric for a sofa, the specifications of a PC, the choice of a seat at an event and so on. That isn’t personalization; it is simply customization.

Nor is it a matter of adding a “Hi, Jordan” greeting at the beginning of an e-mail or a simple “Others who bought this also bought…” recommendation to an online shopping search. Well, then, what is personalization? It’s what happens when you’re searching on a website and are shown items based not just on what others who searched for that term would like, but on what you would like, based on your context, search history and any other items you recently bought. It’s when you phone a call centre and the agent has all the information about you and your history with the company at their fingertips (based on the number you are calling from) and can get right to solving your problem.

Personalization is about creating experiences at scale that get fine-tuned with each successive interaction, empowering customers to get what they want — faster, cheaper or more easily.

At its core, personalization is about speed. Speed in getting to know the customer throughout the customer journey and speed in constantly improving the experience based on that knowledge. Netflix has revolutionized the way we watch television, by building a learning loop for personalization, surfacing massive amounts of content to its base of around 250 million subscribers. For example, it has created more than a million personalized versions of TV-series trailers, using the data on viewers’ reactions to personalize the recommendations viewers see next.

What Netflix has done in movies (and Spotify in music streaming) is now playing out in categories as varied as fashion, grocery, air travel, hospitality, home goods, coffee, home security, insurance and countless others.

Personalization introduces a new way to achieve scale. We have long advocated for building economies of scale, especially in the physical aspects of competition: supply chains, production, distribution. But personalization changes the nature of scale in the customer experience, from the mass production of goods to the mass delivery of one-on-one experiences built on accumulated intelligence.

Personalization at scale and the ongoing cycle of activities that power it create a competitive moat that is very hard to replicate. It is no coincidence that companies across a wide swath of categories, including home improvement (Home Depot), banking (J.P. Morgan Chase), restaurants (Starbucks), grocery (Kroger) and apparel (Nike) have publicly announced that personalized and seamless omnichannel experiences are central to their strategy.

We are now at the point where competitive advantage will be based on a company’s ability to capture, analyze and utilize customer data on a gargantuan scale — and on how it uses that data to understand, shape and optimize the customer’s journey. We estimate that by the end of this decade, personalization leaders across industries will capture almost US$2 trillion in incremental growth.

The five promises of personalization

Done right, personalization goes beyond pushing customers to buy specific products. It’s about making their lives easier. At its most impactful, it can anticipate a customer’s needs even before the customer expresses them. We’re speaking of a proactive process, something altogether different from the sea of targeted marketing messages that get lost in the noise. This process extends well beyond the marketing function; it must be embedded in customer strategy, the digital experience, operations, customer service, employee training, the supply chain and inventory management.

We have distilled the formula for success into Five Promises of Personalization. In the age of AI, every one of these promises must be delivered by bringing together both a human touch and the right technology.

Promise 1: Empower me. This is the overarching promise of personalization, representing your effort to put the customer in the driver’s seat of the relationship. Each customer has a particular set of needs to address, now and in the future, that can be the basis of an enduring relationship. You want to help them achieve a goal, whether it’s inspiring them early in their journey with different choices, educating them about products they are considering, helping them nab a great deal as they consider a purchase or making a return seamless.

To empower your customers, you must first determine which parts of the customer journey are most critical for personalization. Which experiences can you personalize that will help them achieve their desired goal? Too often, companies concentrate on the moment of purchase, with add-to-cart recommendations or on retargeting (finding people in a paid digital channel and serving them relevant ads). To truly live up to this first promise of personalization, think instead about the whole journey and how your use of information builds affinity and cultivates trust.

In the age of generative AI (GenAI), virtual assistants and smart chat interfaces can connect information across multiple systems in ways that previously required implementing expensive integrations or making the customer click through countless menus and multiple websites. New ways to empower the customer are emerging in every domain — from planning a vacation to planning their financial future.

Promise 2: Know me. Being effective at personalization requires knowing the customer. Not the ‘typical’ customer or even a customer segment, but each individual customer. Fulfilling this second promise entails securely organizing, analyzing and synthesizing the data you gather on them and clearly recording how that data can be used.

When the customer sees you using information about them in a positive way, your brand becomes dramatically more relevant. Using information to support the actions a customer will want to take makes their experience simpler, more tailored and fast. Building a detailed, 360-degree view of each customer — with their digital permissions in place — powers all of the intelligence in your personalization technology stack.

Promise 3: Reach me. If a company is not able to connect with the right customer at the right time and deliver the right experience, nothing happens. Therefore, a key promise of personalization is the ability to reach out to the right person, in the right channel, at just the right moment. Fulfilling this promise entails having the insights to know what is relevant to the customer at that moment, based on everything you know about them. The customer must first grant you permission to contact them and use their data in certain ways. From there, you might reach out based on the contextual information you have: the customer’s physical and virtual location, device in use, time of day, weather or some trigger (such as browsing behaviour or an app download).

Both the right timing and access require having enough knowledge and the right intelligence layered on top of that data, to know which customers to contact and when. From a technology perspective, this means having the right targeting intelligence (models and algorithms derived from the data), experiment design and activation (decisions about which customers get which experiences) and next-best-action orchestration (decisions on the timing, sequence and cross-channel coordination of the experiences).

Promise 4: Show me. Giving the customer relevant content is an essential element of personalization. This might be an informational e-mail with tailored text and video, a web or app image, personalized guidance or problem-solving in real time from a call centre rep or a chatbot interaction. Content should be tailored to the information customers expect the company to have on them. Too often, personalized messages — even those from leading digital brands — are generic and uninspiring. Companies need creative content design and operational processes that can generate variations for every type of customer.

Fortunately, new approaches to content delivery and the application of GenAI in content creation are making this much more feasible. Personalization leaders are building content libraries with dynamic templates and modular assets that can be mixed and matched to address different audiences. Instead of duplicating efforts by channel, teams are developing content from the start with omnichannel in mind. GenAI is increasing the productivity of creative teams by making it easier to create additional content variants (e.g. for different segments, languages, personas) and to manage existing content via better tagging.

Promise 5: Delight me. Masterly personalization should feel magical to the customer. Getting the experience just right requires deeply knowing the customer and discovering what delights them over time. One interaction won’t be enough — companies need to set up the processes and ways of working to continuously and rapidly test new ideas to improve the level and accuracy of their personalization.

Learning comes from gaining more information about the customer, as you constantly innovate and try new ways of interacting within each journey, always aiming to delight the customer further. Data that you capture on the outcomes of those interactions then informs the next interaction. And sometimes the lessons involve what not to do.

Companies that fulfill this promise, for example, will stop advertising cars to you when you’re not in the market for a new vehicle and don’t plan to be any time soon. They will also stop sending offers for items you’ve shown no interest in, using the data from your lack of engagement to update their models. And they will stop sending you any repeat purchase offers while you’re sorting through a major issue with customer service.

The drive for continuous improvement is what enables companies to build the competitive moat — of richer information, smarter predictive capability and more receptive customers — by which they can distinguish themselves from competitors. Therefore, delighting customers requires the organizational commitment to relentless innovating, testing, learning and optimizing.

Fortunately, with GenAI and automation, teams can quickly query multiple systems as long as the underlying system is set up to ensure the right data is accessible to all who need it. With every interaction, leading companies capture more feedback about what works, because testing different possibilities is a cornerstone of their approach. When you have the capability to test rapidly, you can constantly try new things, let the technology manage them and focus on pushing the limits of new ideas that will truly delight each customer.

Delivering on the promises: A case study

Starting in 2021, a growing wine and spirits retailer set out to become the Spotify of wine by introducing customers to products they would love. Its story brings to life what it takes to deliver on the five promises of personalization.

• Empower me. First, the company needed to figure out which steps in its customer journey to personalize. Customers were often overwhelmed by the vast number of products to choose from, so the team identified a broad array of marketing actions: recommending items that the customer had never bought before but might like; highlighting new products; and educating customers who preferred a certain varietal about other wines in that category. All of this was in addition to reminding customers to replenish regularly purchased items, suggesting favourites on sale, and offering surprise-and-delight gifts to loyal customers. The company also offered in-store wine classes and tastings.
  
  To accomplish all of this, the personalization team created thousands of messages, with different tones of voice, images, videos and copy to appeal to different types of customers based on where they were in their customer journey. To move quickly, they started with e-mail, but soon expanded to app, web and customer service channels. Content relevant to the customer was synchronized across channels.
• Know me. The company aggregated each customer’s information in a single-view, comprehensive profile to create its customer 360 database. This master repository included all the data needed to maintain a current score for each individual, based on more than a thousand data points. These data points reflected such things as the likelihood of someone being a merlot lover versus a cabernet sauvignon fan; their price sensitivity and propensity to respond to offers; their likelihood of not returning in the next 90 days; and their typical spend and visit frequency. The cloud-based analytics system automatically recalculated the scores every day and fed them back into the database, determining in real time which audience the customer would be classified in for any given marketing action.
• Reach me. Next, the company focused on its intelligence capabilities. The team in charge of personalization — a group of marketers, merchants, graphic designers, copywriters, product managers, data scientists, engineers and marketing technology (martech) experts — built a personalization tech stack they nicknamed ‘Sensei’ (a reference to a wise instructor of Japanese martial arts). Sensei’s set of cloud-based AI models scored the relevance of each piece of content in the content library for every customer based on the customer 360 data, and determined the appropriate piece to send at that time. For example, merlot lovers who were at high risk of leaving might get the merlot discount offer that week. Each new customer interaction generated another 300 data points to add to Sensei to further improve targeting — such as whether the customer clicked on a particular e-mail and how long they spent viewing it and whether they provided feedback via reviews.
  
  Realizing that any marketing action actually consisted of some 20 different decisions, the team embedded 20 different modules in Sensei, designed to optimize a different aspect of each interaction. Take e-mail messages, for example: one module contained different types of themes for the e-mail (say, a reminder or an offer); another, the different types of products to showcase in the message body (based on individual predicted preferences); the next module contained which types of customers would receive the message (e.g. based on hobbies and interests or level of loyalty); another guided when the message should be sent (e.g. times of day and week when that individual is most likely to engage). One showed what tone the copy should take for that particular customer (e.g. educational or fun); and still another stated whether a reminder should be sent a few days later — and so on.
• Show me. The company set up a ‘factory’ to develop its library of content so that it would be able to escalate both the volume and variety of interactions with customers. Working in two-week sprints, the content incubator team — a dozen-plus marketers, copywriters, graphic designers and technologists — filled the library, and within a few months had accumulated 3,000 unique ways to talk about each stock-keeping unit (SKU) in the company’s product catalogue.
  
  Given the number of products and the number of possible content permutations, the team ended up with more ways to talk about its products than it had customers. In this way, it was able to ‘hyper-personalize’ by selecting from its rich content library exactly what would resonate with each customer, based on their unique situation.
• Delight me. The company developed a rapid test-and-learn capability to continuously optimize the experience, based on each interaction. The team ran thousands of experiments to fine-tune the data and algorithms and to determine which types of content worked best in each channel. It created dashboards to track and measure new actions in real time. All of this feedback was used to tailor the next best action.
  
  To further delight customers, they programmed Sensei to continuously learn and improve: 10 per cent of the time, the algorithms would prioritize a message about something the customer might like but hadn’t seen yet, sending the optimal revenue-maximizing communication. The cross-functional team eventually transformed the approach to become the standard way in which the marketing, analytics and digital teams would collaborate going forward.

This retailer’s experience demonstrates that you needn’t be a digital native business to build a world-class personalization capability. Established businesses — even small or midsized ones — can, as well. Digital natives certainly have an advantage: their entire business model is built on digital customer relationships, so they are capturing data from day one. They can also design their technology stack smartly from the start, avoiding the tech debt that plagues many of their established peers.

Most traditional companies today lack the foundational data and technology stack that personalization depends on, which means they must rearchitect their legacy systems. Nonetheless, those that are able to do so will be among the most successful in their categories, substantially accelerating growth and successfully fending off new competitors.

Measuring personalization performance

Over the past decade, our team has worked with hundreds of companies to build personalization capabilities. Since 2016, we have quantified these findings into the BCG Personalization Index, which provides a numeric score from 0 to 100 that reflects a company’s performance in delivering each of the promises of personalization. Several clear insights have emerged from this initiative:

Companies still have far to go to achieve the full potential of personalization. Across all 12 industry segments, the average company scores only 49 on the Personalization Index. The top decile of companies score, on average, only 72. Not surprisingly, the Index leaders are digital natives such as Netflix, Uber, Alibaba and Amazon and early movers such as Starbucks and Sephora.

It is possible to be a personalization leader in any industry. The differences within sectors are more pronounced than those across sectors. Although more highly regulated industries, such as insurance, financial services, and healthcare, tend to score lower, there are nonetheless companies that have handily leapfrogged their competitors.

The Personalization Index findings prove it is possible to build personalization-at-scale capabilities in practically any industry. One example of this is Allianz, the global insurer and asset manager. Although regulations limit insurers’ ability to personalize, companies like Allianz are nevertheless personalizing claim detection and management to make the customer experience more seamless at the moment it most matters, thereby reducing the time it takes for anxious customers to get their claims resolved.

Personalization drives growth and customer satisfaction. On average, companies that score higher on the Personalization Index enjoy faster growth than their lower-scoring peers. This same correlation was observed across sectors and regions, and even held during the pandemic years. The top decile’s six-point annual growth-rate differential compared to the average (and 10-point differential versus the bottom decile) — the size of the competitive advantage from personalization — is consistent with our findings from prior years, and clearly demonstrates that personalization leaders are capturing market share in their respective sectors. The longer-term effect is even more striking: Every one of the personalization leaders we benchmarked is also a market-share leader in its category — and maintained its dominance through recent downturns.

Personalization at scale is much broader and more powerful than anything you can simply place within your marketing function. It takes a full-on corporate effort to assemble the insight, creativity, technology, automation and processes necessary to execute and optimize it.

Importantly, it also takes speed. Personalization leaders are already designing and using every interaction to innovate, rapidly test, learn and empower their customers, making customers’ lives easier with each step. Time is ticking, and the race is on.

This article is adapted from the book Personalized: Customer Strategy in the Age of AI. It originally appeared in the spring 2025 issue of the Rotman Management magazine. If you enjoyed this article, consider subscribing to the magazine or to the Rotman Insights Hub bi-weekly newsletter.