How to Redesign Customer Service for Humans and AI

Service leaders will be most successful in today’s AI-driven world when they create a strong, convincing narrative of why, where, and when humans are essential to advance customer loyalty to a brand. 

For the next five years, AI is unlikely to be preferred by a customer who is anxious, in a hurry, or has an urgent problem. AI survey data and intuition align around the conclusion that these situations require a face-to-face or a chat session, or a video or phone conversation. These indispensable moments are when customer relationships can be sealed for years into the future. Human service reps are essential – they just need to be adapted as you integrate AI into your business. 

In this blog article — the fourth in our series “Agentforce Reinforces the Human and the Humane in Your AI Strategy” — we delve into ways your org can begin transitioning into an agentic enterprise.

What you’ll learn:

• The new customer service job architecture
• As AI replaces tasks, focus on a new age of productivity
• How to start implementing your new roles
• The next three steps for customer service
• The key to a smooth transition: balancing human touch with AI efficiency
  

The new customer service job architecture

So far, many customer experience experts struggle to define the scope of AI agent roles because they don’t have a rigorous definition through which to decide which tasks to remove from humans. We suggest something along the lines of:

For any task currently performed by a human, the resolution of which is achievable by retrieving, updating, or determining an answer using structured data and fixed rules (even where there is context-dependency), the business should prove that the task cannot be done by automation and AI before resorting to humans. In all cases the decision needs to be documented.

This test of whether an AI agent can take over a task from a human is referred to as formal verifiability, which means that a machine can prove that the AI agent will never violate a precisely stated process or guardrail or requirement.

Today, hundreds of straightforward, repetitive, and mundane tasks performed by service representatives answer to this definition. Create your own list of the tasks that service representatives handle that are well within the scope of current AI systems like Agentforce, such as:

• Creating, classifying, summarizing and routing a case
• Searching for knowledge articles
• Order status checks like updates on shipping and delivery, or changing an order 
• Looking up account details, balances, and transaction history  
• Giving a standard response to common, repeatable inquiries about products, product information, services, about return policies, or store operating hours
• Scheduling/canceling/modifying an appointment or a booking, arranging for a field service technician visit
• Processing an item return
• Routing a call
• Outbound surveys
• Coaching cues
• Troubleshooting initial diagnostic steps for technical issues 

Each industry and each type of customer support team will come up with its shortlist to move to AI and automation. (Back to top)

As AI replaces tasks, focus on a new age of productivity

Many customer service jobs are filled with monotonous tasks that create boredom and detachment for employees, which may explain low levels of engagement at work that some companies experience. One could argue that AI offers the perfect opportunity to strip away monotonous and boring work.

But some of the downtime that employees cite as “administrative” and “dirty work” can serve the purpose of providing breaks to recover. As CEOs turn to their CIOs to deliver more efficiencies by substituting human labor with AI and workflow, these CEOs need to ask another question: How can our human employees sustain levels of high productivity when their work consists almost entirely of complex, challenging tasks that AI is unable to perform? 

Service reps move from generalists to specialists

Organizations will need to rethink their work structures and employee management strategies to maintain productivity and employee well-being. This might involve introducing shorter shifts, more frequent breaks, and additional training opportunities to help employees manage the increased complexity and intensity of their work. Some companies are already exploring innovative approaches, such as redesigning workflows to balance challenging tasks with periods of recovery, or creating new roles that use human strengths in areas like strategic thinking and empathy. 

By adopting these strategies, organizations can ensure that their human employees remain engaged, motivated, and equipped to handle the complex tasks that AI can’t perform. (Back to top)

How to start implementing your new service roles

As organizations move forward with AI integration, implementing a human-centered AI strategy will be crucial. This involves not only using AI for efficiency, but also ensuring that human employees are equipped to work alongside AI effectively. Key considerations include:

• Train and upskill employees for high value tasks: Invest in comprehensive training programs that help employees develop skills that complement AI capabilities, such as critical thinking, creativity, and emotional intelligence. This enables them to focus on complex tasks that require human expertise. One example: Salesforce resources like Trailhead — our free online learning platform with practical, role-based AI training — and the Serviceblazer Community on Slack, where support pros share tips, experiences, and best practices for navigating change.
• Focus on employee and organizational health: Implement strategies to prevent employee burnout from high-intensity work, such as regular breaks, wellness initiatives, and flexible work arrangements. This helps maintain a healthy and productive workforce.
• Build new metrics for success: Update performance metrics to reflect the changing nature of work, moving beyond traditional measures like Average Handle Time (AHT) to metrics that capture the value of human-AI collaboration, such as customer satisfaction and employee engagement.
• Look for new opportunities for high-value employees to grow their career into new roles: As AI takes over routine tasks, create new career paths and opportunities for employees to grow into roles that require strategic thinking, creativity, and human connection, such as AI training specialists or customer experience designers. This will also require new hiring practices, higher wages, and an opportunity to revisit long-standing BPO relationships. (Back to top)

The next three steps for customer service

The successful adopters of AI in customer service (like 1-800 Accountant, for instance) will be the ones that define new and emerging roles, inventory what would be easiest and most effective to automate, and plan to upskill the service agents who will handle the more complex interactions with customers.

Here are a few suggestions on your next steps:

1. Over the next 30 days, create a thorough outline of who does what in customer service. Inventory the top contact drivers, label which are the easiest to start with, and define the metrics you hope to achieve, starting with less than 5% of overall service volume. 
2. Over the next 90 days, IT and customer service leaders should create and execute the first Agentforce Service automations. Design and identify the knowledge artifacts, then build, integrate, and test your first agent. 
3. Decide on the support issues that will remain “human first.” These will be the cases that are urgent, anxiety-inducing for the customer, or high in ambiguity. Work on the definition of the new roles. Update the compensation, recruitment, training, and retention policies for these service representatives. (Back to top)

The key to a smooth transition: balancing human touch with AI efficiency

Humans crave connection, and connection with the enterprise forges loyalty and strong word-of-mouth brand amplification. Speed, efficiency, and accuracy are table stakes – empathy, connection, and a human touch are differentiators. 

The temptation will be to use AI agents in these situations as they improve in their ability to be empathetic and to mimic personas. But those organizations that aren’t transparent with customers about their use of AI agents will create a risk of reputational, trust, or financial damage. Customer service requires humans and AI agents, and the leading businesses will be the ones that know exactly how to harness the strengths of both. (Back to top)

This article is part of our series, “Agentforce Reinforces the Human and the Humane in Your AI Strategy.” Check out the others on the Service Cloud blog:

How to Build Humane AI: A Guide for Customer Service Leaders

How to Succeed with AI to Reshape Customer Service Roles

How to Help Your Customer Service Team Thrive — Not Just Survive — in the Age of AI