Guest post by Mark Iafrate, Principal Integrated Marketing at Intercom

As AI becomes a cornerstone of customer service strategies, support leaders face a new landscape for evaluating performance, operationalizing insights, and demonstrating value. This shift requires rethinking traditional metrics, integrating AI-specific performance indicators, and maximizing the broader benefits of AI for both efficiency and customer satisfaction.

In their live discussion, Ben Newell, VP of Product at Geckoboard, and Declan Ivory, VP of Customer Support at Intercom, shared their insights and actionable guidance on adapting their metrics to fit this AI-first world. Below is a recap of the discussion topics and key takeaways. 

You can access a full recording of the webinar here. 

Rethinking Metrics for AI Integration

AI is reshaping the way customer service performance is measured. Traditional metrics like First Contact Resolution (FCR) and Average Handle Time (AHT) are no longer straightforward indicators of success. As AI tools resolve simpler, transactional queries, the remaining workload for human agents often involves complex issues that may take longer to resolve or require multiple interactions. This shift can make it seem as though metrics are "getting worse," but in reality, it reflects a fundamental change in the nature of customer service work.

To measure AI’s impact effectively, support teams need to adopt new metrics tailored to this shift:

  • AI Involvement Rate: Tracks the percentage of interactions AI handles, showcasing its reach and efficiency.
  • Resolution Rate: Measures how many queries AI resolves without human involvement, reflecting the quality of its performance.
  • Inferred Metrics: Tools like Intercom’s Customer Experience Score use AI to evaluate interactions, even in cases where customers do not provide direct feedback.

These new metrics help teams understand not only how AI contributes to efficiency but also how it shapes the overall customer experience. As these tools mature, leaders must analyze both human and AI performance holistically to provide a complete picture of success.

Operationalizing Metrics to Drive Performance

Metrics are only valuable when they drive actionable insights. For human support, real-time metrics can be transformative, allowing teams to identify and address issues as they arise. For example, a team might shift resources to handle an increase in phone queue volume or adjust workflows to improve response times. Real-time dashboards empower teams to self-organize and solve problems collaboratively.

"Real-time metrics empower teams to address bottlenecks immediately, fostering ownership and agility." —Ben Newell, VP of Product at Geckoboard

AI metrics, on the other hand, require a different approach. Because AI performance trends often emerge over time, leaders need to analyze data at a slower cadence, identifying patterns and opportunities to improve content, workflows, and guardrails.

Cross-functional collaboration is essential in operationalizing metrics. AI's success depends on input from multiple teams—R&D for feature optimizations, sales for aligning customer expectations, and customer success for ensuring that AI-driven processes deliver value.

"AI is a teammate—measuring its CSAT alongside humans ensures consistency in customer experience." —Declan Ivory, VP of Customer Support at Intercom

To ensure the right insights reach the right teams, support organizations should establish regular feedback loops, leveraging dashboards and reporting tools to share updates with stakeholders. By aligning metrics across teams, organizations can ensure AI and human support work seamlessly together.

Measuring ROI: Beyond Efficiency Gains

AI’s most apparent contribution to customer service is efficiency. Automating routine tasks reduces the workload on human agents, helping organizations manage growing customer demand without a proportional increase in headcount. However, focusing solely on cost savings misses the broader benefits AI can provide.

By integrating AI, companies can elevate their service offerings in ways that were previously unattainable:

  • Enhanced Service Levels: AI enables capabilities like 24/7 multilingual support and proactive outreach, ensuring customers receive assistance whenever and however they need it.
  • Proactive and Consultative Support: Freed from routine tasks, human agents can focus on helping customers get more value from products, identify new opportunities, and resolve complex issues with a personal touch.

The ROI of AI extends beyond efficiency. By improving the customer experience and deepening customer relationships, AI drives long-term value, including higher retention rates, increased customer loyalty, and reduced churn.

Conclusion

AI is transforming customer service by changing the way organizations measure success, operationalize insights, and deliver value. By adopting new metrics, creating effective feedback loops, and focusing on both efficiency and customer experience, support teams can maximize the benefits of AI while preparing for its continued evolution. Organizations that embrace this shift will be well-positioned to meet rising customer expectations and build stronger, more loyal relationships.