Resourcesicon
Agentic AI and the Evolution of CMA Program Management
Woman looking at computer screen with several mini robots floating near her, representing agentic AI.

Agentic AI and the Evolution of CMA Program Management

For decades, Customer Marketing & Advocacy (CMA) has been the unsung hero of B2B growth. It’s the engine behind those success stories that inspire prospects, the whispered endorsement that tips a deal in your favor, and the subtle force that makes your demand gen campaigns land with credibility.

Yet, for most organizations, CMA has been chronically underfunded, short-staffed, and—let’s be honest—often underappreciated. To too many executives, CMA looks like a cost center. Something “nice to have.” Something expendable when budgets get tight.

For those of us who have lived and breathed CMA, this is maddening. We’ve seen the magic: customers who speak at your events, advise buyers based on real-world experience, and turn your sales cycle into a trust-building exercise rather than a cold pitch. When an executive *gets it*, the possibilities are endless. Unfortunately, there just aren’t enough of those executives.

Enter Agentic AI: A Door Cracks Open

If CMA has struggled to get attention from the c-suite, the current AI frenzy may just change the game. Boards are pressuring leadership teams to turn AI into ROI—fast. Executives are leaning on managers to show tangible progress—if not results. Suddenly, there’s appetite for experimentation, and the once-overlooked CMA function has a shot at the spotlight.

This is our moment. But it comes with a twist.

Agentic AI—the emerging class of AI tools that can act autonomously, interact conversationally, and adapt to situations—is unlike the automation of the past. Automation was predictable, procedural, and safe. Agentic AI is…well, a little “schizophrenic.” It hallucinates. It can produce five different answers to the same question. In domains like CMA, where trust and relationships are currency, that’s not a small risk.

Still, in this rare season of executive curiosity (and tolerance for AI’s warts), CMA leaders can step into a new era—if they balance ambition with caution.

A Glimpse at the CMA Program Manager of Tomorrow

Let’s imagine how the role of a CMA program manager will evolve as agents become capable teammates rather than just tools. Take for example, the ongoing chore of discovering customers ripe for advocacy.

Example: Advocate Recruitment on Autopilot (Sort Of)

Recruiting advocates today is notoriously time-consuming. It depends on your relationships with sales reps, CSMs, product marketers, and executives—and that’s before the first outreach email goes out.

In a mature agent-driven environment, much of that manual work could be offloaded. Imagine an agent that:

  • Scans your CRM and other relevant data sources to identify customers who fit your Ideal Advocate Profile (IAP), and express advocate-ready sentiment.
  • Cross-checks for red flags: open support cases, pending renewals, or poor survey scores.
  • Reaches out to prospects in a personalized way, gauging interest and capturing profile info.
  • Onboards advocates, triggers welcome kits, and gracefully escalates to you if something feels off.

It sounds magical—and in some ways, it will be. But designing this magic requires deep CMA expertise, in addition to competent interaction design. Without it, the agent might create social faux pas that damage relationships instead of nurturing them. AI providers acknowledge that humans are often better suited for dealing with customers in marketing, sales, and customer success. Human employees are essential for managing and resolving sensitive matters with customers, and what’s more sensitive than asking a customer to put her reputation on the line for your company?

From Manual Labor to Strategic Leadership

As agents handle more operational tasks—candidate identification, outreach, data collection—the CMA program manager’s role shifts. Tomorrow’s CMA leader will:

  • Act as an agent orchestrator and tuner, refining prompts, rules, and escalation points.
  • Monitor agent outputs, spotting errors before they reach the customer.
  • Spend more time on program strategy, optimizing IAPs to align with evolving business priorities.

Think of it as moving from an assembly-line role to a command-center role. Tools like Salesforce’s emerging Agentforce Command Center will make it easier to track performance and refine agents. Because agents need direction, it will be incumbent on program managers to invest more in strategic planning and program design than in the pre-AI world. The consequences of not doing so will be dramatic, visible and fast within an organization.

Recognizing the Boundaries of Agentic AI

No matter how smart the agent, there will always be gaps in the data. For example:

  • A buyer wants to speak with a customer contact who led the product implementation—but your database doesn’t track that information.
  • Your ideal advocate contact at the perfect account is unexpectedly unavailable—whether they’ve just gone on vacation, maternity leave, taken a sabbatical, or left the company entirely. If there isn’t up-to-date availability data, an agent can’t (and shouldn’t!) fabricate that information out of thin air.
  • If the ideal advocate isn’t available, the next best advocate contact choice is not identified in a database. This requires fuzzy logic that an agent would likely get wrong.

These situational nuances require human judgment. A savvy program manager will design workflows where agents flag these scenarios for human intervention, preserving the buyer experience—and the deal.

The New CMA Superpower

For the first time, CMA program managers have a shot at scaling their programs without scaling their headcount. Agents can become true virtual team members, freeing program managers to do what they’ve always wanted to do:

  • Build richer advocate relationships
  • Act as customer advocate consultants across the enterprise
  • Align more closely with company growth goals
  • Drive measurable influence on revenue

But this will only work if we combine deep CMA expertise with smart agent design. Without that foundation, organizations risk over-rotating on AI, frustrating customers, and undermining trust—the very currency of advocacy.

At Point of Reference, we’ve poured over 20 years of CMA experience into our agent design principles. Technology may power the future, but empathy and context will be our north star.

The CMA program manager of the future won’t just manage relationships—they’ll manage the agents who manage the relationships, elevating both the human and digital sides of advocacy. And for those ready to embrace this shift, the sky’s finally the limit. For more on the role of AI in Customer Marketing, check out this podcast with Sunny Manivannan.

-