This writer outlines what could be undermining customer experience despite advanced digitalisation, and offers checklists for achieving (positive) resolution-focused operations.

For many organisations in the Asia Pacific region (APAC), customer experience (CX) is a critical differentiator. Customers interact across multiple channels, expect quick and accurate responses, and are quick to switch providers when issues go unresolved.

In that context, resolving a customer’s problem is central to keeping their business.

Then, as organisations look to improve customer service, AI is offering new possibilities. Adoption across the region is already accelerating. Yet investment and outcomes are not the same.

If organisations already have the technology to respond faster, why does solving the customer’s problem still feel so difficult?

Carlos Quaderi, Head of Asia Pacific, Zoom

Disconnected tools, disconnected outcomes

Many organisations in APAC do not lack digital tools. They run collaboration platforms, CRMs, messaging apps, AI assistants, contact center systems, and workflow tools, often all at once.

The issue is that these systems still largely operate in isolation: more technology, but also more fragmentation, where conversations start in one channel, move to another, and lose context along the way.

Customers may never see the internal lines between front office, contact center, and back office teams, but they experience every broken hand-off. When context does not travel with the interaction, the next employee or AI agent effectively starts from zero. Staff spend more time switching between systems and chasing information than helping the customer.

Many legacy self-service tools were built as keyword-matching systems layered on existing infrastructure. When a conversation moves beyond the script, the customer is handed off without the context already established. For example: A customer calls about a billing dispute. The agent responds in one system, escalates in another, and logs notes somewhere else. Two days later, the customer follows up, only to find the context never carried over. The interaction was completed; the issue was not.

Such fragmentation starts with data. Customer information is often scattered across systems and teams with no shared view. Layering bolt-on AI into that environment compounds the problem: AI is only as reliable as the data it can see. When data is incomplete, bot conversations degrade and agent recommendations become less reliable.

For APAC businesses, particularly small and medium-sized enterprises with leaner teams, these inefficiencies are harder to absorb. The question is no longer whether organisations have adopted AI, but whether their systems can share context and carry workflows forward without creating more friction.

This is why interoperability matters. Businesses increasingly want open environments that allow different systems, workflows, and AI models to work together, rather than forcing everything into a single closed ecosystem.

Checklist for leaders assessing their CX stack

  • ✓ Map the top 3 – 5 customer contact journeys end to end (channels, systems, teams, hand-offs).
  • ✓ Identify where context is lost at each hand-off (missing transcripts, case IDs, status, prior actions).
  • ✓ Define a minimal “context envelope” that must travel with every case (e.g., customer ID, issue category, last interactions, current status, next SLA).
  • ✓ Treat missing context at hand-off as a defect, not an acceptable quirk.
  • ✓ Prioritize integration work on the journeys that generate the most repeat contacts and rework.

Checklist for organisations with fragmented tools

  • ✓ Require that every escalation or channel switch passes the full interaction history and case status.
  • ✓ Enforce use of a single case ID per issue across all systems, even if integrations are limited.
  • ✓ Log where agents must manually re-enter data or re-gather customer information; use this to target integration or workflow changes.
  • ✓ For high-impact journeys, wire up automated context passing between the most critical systems first (e.g., contact center ↔ CRM ↔ billing).
  • ✓ Maintain a simple, shared view of “critical unresolved” cases (owner, last action, next step) if full integration is not yet possible.

Why positive case resolution matters

For years, CX has often been measured through speed-based metrics: average handling time, ticket deflection, and containment. Operationally, those indicators still matter, but they do not show whether the customer’s problem was actually solved.

A hand-off is not a resolution, and a closed ticket is not always a solved problem. If the customer has to return, repeat information, or follow up through another channel, the business may have recorded an interaction as complete while the customer still sees the issue as unresolved.

In this environment, businesses are judged less by how quickly interactions happen and more by how completely needs are addressed across the journey.

Consider an insurance advisor handling policy renewals and claims follow-ups. Under a speed-first model, the focus is on closing tickets and keeping handling times low. But if context does not carry across channels and teams, the same customer ends up repeating information at the next interaction. Efficient on paper; broken in practice.

Under a resolution-focused model, priorities shift. Did the issue stay resolved? Did context carry forward without repetition? Did the interaction reduce effort rather than create more? In APAC, this matters because CX teams work across multiple languages, markets, and service expectations. Every break in context adds cost to the operation and the experience.

Checklist for teams deploying or using AI agents

  • ✓ Define “resolved” for each AI-handled use case before go-live (what outcome counts, what triggers human escalation).
  • ✓ Ensure AI-to-human hand-offs include the complete transcript, data looked up, and actions already attempted.
  • ✓ Monitor resolution, not just containment or deflection:
    • % of AI-handled cases that do not generate a repeat contact for the same issue within a set period
    • % that require human follow-up within 24–72 hours
  • ✓ Regularly sample AI responses and verify they reflect the latest customer data, policies, and edge cases.
  • ✓ Correct data gaps at source rather than only tweaking prompts or flows.

From systems of record to systems of action

If positive case resolution is the benchmark, businesses need more than systems that store information. They need systems that help work move forward. Systems of record log interactions and preserve notes. Yet CX rarely breaks down because data is missing; it breaks down because execution fragments once the interaction ends.

After each client conversation, there are notes to write, follow-ups to schedule, approvals to escalate, and next steps to hand off across teams. In many organisations, this still happens manually, across disconnected systems, at the agent’s initiative. The conversation happened; the completion did not.

Systems of action are designed to carry work from conversation to completion, rather than simply automate individual tasks. When interaction history, customer signals, and next steps move together across channels and teams, organisations are better placed to resolve needs completely, reduce repetition, and avoid the dropped threads that erode trust.

Rolling out an AI agent is not the finish line; it is the starting line. The harder work is monitoring whether the AI continues to resolve what it was designed to resolve, and correcting course when it does not.