On Customer Support

The tech industry has a habit of chasing silver bullets for customer support: chatbots, self-service portals, omnichannel communication. Each new tool promises to revolutionize the way companies serve their users. Yet the results are often underwhelming. Why? Because the problem isn’t about the tools—it’s about the approach.

Customer support typically deals with three categories of users:

1. High-need users: Those who struggle with technology and require substantial guidance.
2. Middle-of-the-pack users: The majority, facing routine issues that are easy to resolve.
3. Edge-case users: Experienced users dealing with unusual or complex problems.

Metrics-Based Myopia Most companies focus their efforts on optimizing support for the middle group. Metrics like average handle time (AHT) or first contact resolution (FCR) favor this approach because these users are the easiest to satisfy. But this focus leaves the two extremes underserved.

• High-need users feel alienated by self-service tools they can’t navigate.
• Edge-case users become frustrated by generic scripts that waste their time.

This misalignment creates vocal dissatisfaction and bad experiences that ripple through reviews, social media, and churn rates.

The Edge Case Failure: A Real-World Example

We once worked with a vendor whose product faced a critical issue. With strong monitoring tools, we quickly identified the problem and reached out to their support team, confident we could guide them toward a solution.

Instead, we were met with scripted responses that assumed user error. Despite having a dedicated account manager, we spent nine hours exhausting their standard playbook before convincing them to look at the issue we had already diagnosed. Once they did, it was fixed in minutes.

What could have been a 30-minute outage turned into an all-day ordeal, wasting resources on both sides and causing frustration for everyone involved—all because of an inflexible system designed for average issues, not edge cases.

Reframing the Solution: Smarter Customer Support With AI

The solution to better customer support isn’t about adding more channels or tools—it’s about smarter prioritization and targeting. AI offers a way to bridge this gap:

1. Classify Support Requests: AI can analyze the nature of incoming support tickets to classify them into categories:
2. Route Intelligently: Once classified, issues can be directed to the appropriate team. Routine problems go to chatbots or entry-level support agents, while high-priority or edge-case issues are escalated to specialists who can skip the script and act quickly.
3. Personalize Support: AI can leverage customer data to tailor responses and identify users who might struggle with self-service solutions. Offering live assistance to high-need users increases satisfaction without wasting resources on unnecessary escalations.
4. Focus on Results, Not Process: Move away from rigid adherence to scripts. Allow experienced agents the flexibility to bypass unnecessary steps and focus on resolving the issue efficiently.

Building the Right Team

One valid concern is how to train skilled support agents who can go off script. This requires rethinking how companies structure and empower their support teams:

1. Build Strong Relationships Between Support and Engineering: Support teams often operate in silos, disconnected from product and engineering. Breaking down these barriers allows agents to better understand the product and escalate issues effectively.
2. Evaluate Your Support Structure:
3. Mentorship and Incident Response: When engineering teams handle incident response and speak directly to customers, it sets a high standard. Customers value transparency and direct answers. Pairing engineers with support teams can build this kind of trust and foster learning.
4. Develop Feedback Loops: Encourage a culture where support teams share insights from real-world issues with engineering and product teams. This collaboration improves both customer experience and product quality.

Why It Matters

Customers don’t care about platitudes or promises—they care about results. Failing to address the extremes—those who struggle and those stuck in complex situations—alienates your most vulnerable and your most vocal customers.

With AI as a tool for classification and routing, and with skilled support teams empowered to act, companies can break free from the inefficiencies of generic support. The focus should be on resolving problems at the source, not just placating users with scripted interactions.

Takeaway: Stop optimizing for the average. Build systems that can handle exceptional cases with the same care and efficiency as routine ones. Customers at the extremes are your biggest critics—but when served well, they can become your loudest advocates. Make them a priority.