The Rise of Agentic AI: The Next Evolution of Personalization

Boston Consulting Group is flipping the script on digital marketing, helping brands personalize their approach and empowering consumers all at once.
Image may contain Body Part Hand Person and Game

Today, we’re well accustomed to eerily personal digital marketing. If you're searching for a top-line lawn mower for your dad’s upcoming birthday, chances are a couple quick internet searches will work just as well as actually going shopping. But as the idea that your digital domains might know you better than you know yourself has evolved from bewildering to routine, a new kind of question has emerged: If this is our new normal, why aren't the outputs of personalization ... better?

Here's the reality: Despite the advances of personalization, there’s still a lot of guesswork involved. Cookies track website visits but can’t capture the intent of the visitor (like shopping for a lawn mower versus needing a whole catalogue of yard tools). Retargeting helps marketers put viewed products back in front of a customer, but this blunt method can be as exasperating as it is enticing (you live in an apartment, after all!).

But what if, instead of seeking customers through ads, brands connected to consumers in a more meaningful way—through personalized, helpful agents, ultimately leading to more positive results for brands and consumers alike? What if these brands could understand your intent, and deliver on it, holistically?

This—the part where digital recommendations flip from sometimes helpful, usually annoying, to the opposite—is the untapped potential of personalization. And with a new solution centered on agentic artificial intelligence (agentic AI), the experts at Boston Consulting Group (BCG) are helping businesses realize it as they outline in their new book, Personalized: Customer Strategy in the Age of AI (published by Harvard Business Review).

A Closer Look at Agentic AI

“Agentic AI is about allowing a customer to pull the solution to their problem instead of a brand pushing a solution toward them,” says Mark Abraham, managing director and senior partner at Boston Consulting Group and the group’s global personalization leader.

Imagine a completely personalized agent that goes beyond the responses of a familiar generative AI (gen AI) chatbot. This helper would take action on a consumer’s behalf and help them achieve a goal. Someone remodeling their home, for instance, could interface with the virtual agent, which could scour the web for the most effective methods and materials, talk to stores, order products, find (or even generate) self-help videos, and make recommendations according to the user's preferences.

“It's really about offering a value proposition to empower somebody,” says David Edelman, a senior advisor at Boston Consulting Group who has been working in personalization since 1989. “You have to want to do this in a way that's going to actually add value for the customer, not just manipulate them for you to eke out a few more basis points in lifetime value,” he says.

It’s this next level of support, say Abraham and Edelman, that will truly fulfill the promises of personalization in AI-powered marketing. And when it’s done well, everyone wins: Businesses save on costs, while consumers more easily find exactly what they need.

Incorporating Agentic AI Into the Customer Journey

That data comes from the Personalization Index: a tool the two, along with their team at BCG, have developed to rate how well companies are delivering on those promises, and where they have opportunities to leverage agentic AI to improve.

Using the index, BCG has scored hundreds of companies around the globe on about 100 different metrics, offering specific insights into their performance and comparing them to industry peers. (And they’ve discovered proof that their methods are effective: Companies that the framework identifies as personalization leaders—that’s only about 10 percent—are growing revenue an average of 10 points faster than the rest.)

“You can’t improve what you can't measure,” says Abraham. “Oftentimes, the first discussions I have with companies are just getting them aligned on what they mean by personalization. And then actually coming up with an objective way to measure their progress year over year, relative to their peers.”

The index examines a two-part question: Does an organization have the ability to deliver an empowering experience—and if so, are they actually delivering it? “These organizations often have the capabilities; they poured money into marketing technology, they’re spending on AI,” says Edelman. “But strategically, they're not focused on delivering that evolved experience. They haven't prioritized it.”

Fulfilling the Five Promises of Personalization

So where should organizations start as they seek to implement a truly strategic approach to personalization? To help these organizations refocus their thinking accordingly, Edelman and Abraham have outlined a framework they call the Five Promises. Each is an area where today’s customer is expecting brands to deliver—and where agentic AI has the power to raise the bar.

  1. Empower me. “The promise of empowering is really what this is all about,” says Edelman. “It’s not just about traditional marketing methods, like putting someone's name into an algorithm or retargeting them.” As an example, he highlights an industrial distributions client that delivers an agentic AI solution to roofing companies. The solution enables roofers to input materials, weather, and zoning details, then provides a range of options that meet their specifications and delivery needs. From there, the tool actually places the order for those companies—leading to sales for the distributor and a more seamless experience for the customer.
  2. Know me. The best agentic AI developers are building clear feedback loops into virtual assistants so that the more customers use them, the better the results become. “For you to have more control, the tool needs to learn: Do you want good, better, best?” says Edelman. “At the beginning, agentic AI is going to be probing for that information and giving you an option to share more feedback—so that it can get smarter in how it gets things done for you.”
  3. Reach me. Companies need to think carefully about when AI-driven targeting is still appropriate versus when to rely on customer-initiated interactions. An easy example of what it looks like to strike this balance is through an app—a tool that can send push notifications, but is essentially a convenient and controllable space for the user. “It reaches you when you need it, because it's right there,” Edelman says.
  4. Show me. Modern shoppers love seeing what they’re ordering, whether it’s a picture of the lunch order they’re about to make or a virtual look at how the jacket they’re eyeing might fit their body type. To meet this demand, brands need to build out their content libraries by using a variety of methods—including gen AI. “You’re going to see a lot more multimedia content generated,” says Abraham. “Getting things like hands and human forms right is not quite there yet, but we're on the cusp of progress.”
  5. Delight me. As agentic AI assistants evolve to tune into their users’ preferences, they should be able to surprise and delight users with their responses and recommendations. “These virtual assistants, by nature of the AI they’re tapping into, will be more and more personalized,” Abraham explains. Eventually, the virtual agent will learn a customer’s personal tastes, budgets, habits, and more—enabling them to offer increasingly targeted, accurate, and satisfying results.

With only 10 percent of companies hitting the mark in their personalization efforts, the shift to this new future of personalization surely won’t arrive all at once. Some industries, Abraham and Edelman note, are naturally behind others. Healthcare, for example, may be slower to adopt patient-facing tools given the industry’s stringent regulatory standards and inherently low risk tolerance.

**Related: Discover the BCG tool that’s transforming AI testing and evaluation **

Yet on the other hand, we are already seeing gen AI tools delivering virtual assistance to millions of clinicians. Edelman and Abraham go on to cite countless customers who have made strides: from a major bank’s virtual assistant that helps customers achieve financial wellness, to the auto manufacturer whose AI tool allows customers to select and review options to assemble the best car for their needs. Across industries, change is in motion.

This is all just the start, they say. As agentic AI technology is refined and companies take the steps to embrace it, consumer experience will continue to improve—and above all, that shift is centered on empowerment. As Abraham reiterates, “This is the fundamental thing we're talking about: giving people the opportunity to find their own solutions, instead of waiting for the brand to give it to them.”