AI's new role in customer success is to automate non-value-adding tasks so CSMs can focus on more strategic, high-value interactions.
Bobby Cooper, Principal Architect at Pilr CX, explains why CX leaders should design systems that integrate AI agents alongside human employees.
The four-pillar framework for AI integration is already yielding significant cost reductions and improved customer satisfaction in reactive support.
Most conversations about the future of Customer Success are dominated by a single question: which jobs will AI replace? The fear of replacement is palpable for employees across industries. But in sectors like CX, where human relationships have long been the gold standard, it's particularly acute. However, some experts say this narrow focus on replacement misses the bigger picture. Instead of subtraction, many see transformation as a ripe opportunity to fundamentally re-architect traditional value pipelines.
We spoke with Bobby Cooper, a veteran CS executive and the Principal Architect at Pilr CX, a company building modern CX architecture that scales retention. After more than 13 years leading post-sale teams at high-growth SaaS companies, Cooper recently launched his own consultancy with an exclusive focus on AI transformation in the customer experience. From his perspective, the debate over whether AI will take CSM jobs is already settled. The more important question is how to manage these new, AI colleagues.
A new kind of org chart: As the lines between human and machine labor continue to blur, the most effective leaders will be the ones to stop managing people and start designing systems. "Soon, org charts will be filled with both humans and AI agents. But the path to get there will be a deliberate one." To drive the point, Cooper likened the process to "creating a job description for an AI agent with specific goals and metrics to hold it accountable." Also important is performance management, Cooper said. "AI messes up. It's not perfect, so anticipate needing to correct the workflows and automations."
The ownership mandate: A clear chain of command is also necessary in this new AI era. Ideally, one that prevents the chaos of disconnected tools and broken workflows. "They need to be under the singular department head's umbrella," Cooper argued, "that means the VP of CS owns the AI agents, just like they own the CSMs. Performance management must be equal and under the same umbrella."
Next, Cooper laid out a clear framework to break the customer journey up into four distinct areas, each offering different potential for AI engagement. In his view, this is the strategic map leaders should follow to move from 'vibe-buying' or haphazardly buying tools to intentionally architecting an intelligent system.
A four-pillar CS framework: Cooper examines the entire customer experience during engagements. "I focus on the reactive or inbound experience, the proactive or outbound, value-driving experience, customer onboarding, and customer operations, the technical systems layer underpinning the whole thing." Interestingly, Cooper said, the level of AI engagement varies across these four areas.
But the quickest and most dramatic wins are often found in the reactive and operational pillars. For instance, in written communication, Cooper said, AI can already outperform humans in many ways.
Domain expertise on demand: For example, Cooper referenced a scenario that's all too common in sales. "One company I worked with sold dental software to multiple personas: the dentist, the hygienist, the front desk manager," Cooper recalled. At the time, "it was hard for entry-level customer service folks to speak fluent domain language with doctors. But an AI can do that quickly. Now, learning how to speak like a doctor or a hygienist and knowing all the ins and outs, from the names of different pathologies to the technical aspects of an office manager, is relatively easy."
The economic case: Also important, Cooper noted, is the financial impact. "Another company had a fully burdened cost of about $6 per message. We reduced that to about $2.50 per interaction once they started an outsourced team. After that, we launched a generative AI chat system that dropped those costs down to about 35 cents per interaction. We also increased customer satisfaction scores, decreased wait times, and increased the rate and clarity of product feedback." Done right, Cooper said, the outcomes can be significant.
The same logic applies to customer operations. Here, AI agents can automate the manual tasks currently consuming hundreds of hours of work. With this type of efficiency gain, Cooper said, leaner teams are possible. Eventually, it could reshape the very nature of CSM roles.
Automating the mundane: At one company, "the CSMs were spending about 200 hours a month on one particular manual task," Cooper said. "Once a customer reached 30 days, the CSM would log into a system, look at their utilization data, and categorize them as 'low,' 'mid,' or 'high' usage. Then, they would select a manual email template, copy and paste it, and send it off. So, we operationalized it, which took about a week."
Elevating the human role: Cooper is quick to clarify that agentic AI is not about eliminating the human element, but elevating it. "I am not saying AI should replace all CSMs," he asserted. "I'm saying AI is replacing all of the non-value-adding activities so that CSMs can now spend 100% of their time interfacing with customers in high-value conversations."
Yet, Cooper cautioned, this is not a one-size-fits-all solution. Areas like customer onboarding, which require deep empathy and human-led change management, are still challenging to automate. Even people who love buying software often hate implementing it, and a human touchpoint is usually critical for navigating this friction. In this new mandate for leaders, the future doesn't belong to the companies that buy AI. It belongs to those who architect with it. "Create your customer experience with intent. How we deliver that value is a blank canvas right now. The possibilities are endless for what we might achieve."
