The conversation with Rob Walker, GM of 1:1 Customer Engagement at Pega, begins with a blunt truth: if we work the same way next year, we’re already behind. That challenge frames a deeper discussion about what “empathy” really means in enterprise systems – and how artificial intelligence is forcing companies to operationalize it.
From Empathy as Value to Empathy as System
In Walker’s world, empathy isn’t a slogan or a soft skill – it’s a system property. Pega’s one-to-one engagement model takes the old notion of personalization and makes it dynamic, adaptive, and ethical. Instead of deciding what the business wants to push, the model asks: what’s right for this customer in this moment?
That orientation changes everything. Decisions are sequenced like a hierarchy of needs: fix hardship first, solve service issues next, and only then – if context warrants – extend an offer.
The strategy bends to the customer’s lived reality, not the other way around. When done well, empathy becomes a performance driver: relevance builds trust, trust compounds into lifetime value, and both sides win.
The Content Bottleneck Breaks
Real empathy, though, is hard work. It requires an awareness of tone, timing, and context that humans can’t sustain at industrial scale. Historically, the roadblock has been content breadth – you can’t have a truly personalized conversation with only a handful of pre-written messages.
This is where generative AI changes the economics. Large models can expand message libraries, rewrite formats, and generate explanations at machine speed while remaining on brand and compliant.
In effect, AI gives empathy a palette. Walker points to studies showing that, in narrow contexts, large models demonstrate better bedside manner than humans because they can pause, mirror sentiment, and avoid fatigue. Scaled across millions of interactions, this capability makes empathy measurable and repeatable rather than aspirational.
The Hype Question
Every serious AI discussion circles back to hype. Walker’s take is nuanced: yes, there’s noise but the underlying shift is real.
One of the clearest examples is redecisioning. Each time a customer hesitates, changes channel, or signals frustration, the system recalculates context and re-selects the next best action in real time. That’s not analytics; it’s an operating model.
Redecisioning requires guardrails, and here Pega’s T-Switch framework becomes critical. It codifies transparency and model-use policy by decision type. A loan approval might demand full explainability and human oversight, while a background-color test can safely run with lighter governance.
By embedding these controls in production systems – not just policy documents – trust becomes enforceable. It’s a practical answer to AI’s credibility problem: fairness, safety, and relevance are built into the machinery.
When the Customer Has an Agent
Looking further out, Walker sees another tectonic shift: customer advocacy agents. These next-generation AI assistants will act on behalf of individuals – shopping, comparing, negotiating, and resolving service issues without human friction. They won’t care about glossy ads or brand narratives; they’ll optimize for user preference, budget, and reliability. That could flatten marketing’s emotional theater and push competition toward transparency, price, and fit.
The implications cut both ways. Empowered customers will gain a tireless negotiator, leveling access for people who lack time or fluency to advocate for themselves. But privacy and interoperability pressures will intensify. Brands will have to publish machine-readable value; clarity of data, clear consent rules, and trustworthy integration, because humans won’t be the only ones listening.
Inside the Enterprise: A Different Kind of Creativity
Internally, generative tools are already rewriting how engagement strategies come to life. Tasks that once took months – mapping models, writing eligibility rules, authoring hundreds of content variants – can now be bootstrapped in hours. Pega’s Customer Engagement Blueprint, for instance, can ingest a company’s site and creative briefs, then propose a governed, testable engagement strategy almost instantly.
This speed doesn’t replace marketers; it repurposes them. The focus shifts from producing collateral to designing orchestration systems – setting policy, defining tone, supervising guardrails, and interpreting outcomes.
Data scientists, in turn, move from manually crafting features to monitoring drift and bias, ensuring performance remains explainable and fair. The creative act becomes governance itself: curating the human values that shape machine decisions.
The Real Risk: Working Like It’s 1979
The danger isn’t that AI will overrun marketing, it’s that organizations will cling to obsolete habits: batch campaigns, static segments, manual queries, and week-long waits for analysis. Modern engagement operates at the speed of signal. Waiting for quarterly reports is like using a sundial in a world of atomic clocks.
Education becomes the new moat. Many seasoned marketers simply don’t realize what’s possible: live redecisioning in every channel, adaptive guardrails that enforce ethics without friction, and generative content that scales empathy instead of undermining it. Lower the barrier to learn through transparent demos, open tools, and hands-on blueprints, and adoption accelerates. Trust follows.
Trust as the Compound Interest of Empathy
Walker’s final point lands where it began: trust is the compounding return on empathy done right. Customers reward relevance with attention and data; brands reinvest that data into better service. Systems that sense and adapt in real time aren’t just efficient – they’re humane at scale.
The coming year will test who’s serious about that promise. Companies that embed empathy into their operating model – not just their messaging – will lead the next wave of customer engagement. Everyone else will still be talking about “personalization” while the market moves on.
