[Voice AI Market Forecast Report] Call Center Voice AI Market Expected to Reach Approximately 525 Billion Yen in 5 Years

## Voice AI Market Forecast Report

株式会社Verbex, which develops and provides voice-enabled AI platforms, has announced its independent estimate of the voice AI market in the domestic call center industry.

Faced with chronic labor shortages and declining response rates during peak call volumes, the call center industry is under pressure. Verbex predicts that over the next five years, approximately 25% of the total addressable market for voice AI in call centers (approximately 2.1 trillion yen*) will be substituted by voice AI operations, totaling approximately 525 billion yen.

*Calculated by excluding non-voice channels (35% estimated by Verbex) from the 1.7 trillion yen outsourcing market (2025 CCAJ Contact Center Survey) and adding the 1.5 trillion yen in-house market (Verbex mid-term management plan 2025), excluding non-voice channels (30% estimated).

### Transforming Call Center Operations with Voice AI

Rather than automating everything, the standard for the future will be a division of labor between AI and humans based on task routine and the need for emotional consideration. Verbex categorizes the suitability of voice AI in call centers into three phases.

#### 1. Early Adoption Phase: AI-Driven Tasks (Routine and Well-Defined)

Tasks where the goal is clear and customer requests can be easily categorized are most suitable for voice AI. Examples include initial intake, requirement gathering, routing to departments, FAQ responses (e.g., business hours, delivery status, billing inquiries), appointment changes/cancellations, and basic cancellation requests.

E-commerce order processing is particularly promising as it faces high economic impact from peak-time call congestion.

#### 2. Gradual Expansion Phase: Hybrid AI-Human Model (Risk Management Included)

This involves broadening scope gradually. Examples include procedures requiring identity verification, closing cancellation requests, special processing, and final orders for high-value items.

A hybrid model where AI handles intake and data collection and immediately transfers to an operator when exceptions occur is realistic. Large-scale IVR systems in major enterprises often require phased replacement.

#### 3. Human-Driven Phase: High-Value Tasks (Emotion, Judgment, and Responsibility Required)

Human roles are essential when customers seek empathy beyond mere information. Examples include handling complaints, emotional consultation, critical negotiations, and inquiries with excessive exceptions. The most realistic evolution is for AI to handle the entry point, allowing humans to focus on high-value interactions.

### The Near Future of Call Centers

* Eradicating "Can't get through" stress: 24/7/365, zero wait time, instant response.
* Operators as "Highly Skilled Professionals": Routines to AI, humans focus on emotional care and negotiations.
* Phone data as "Management Assets": Inquiries automatically analyzed and structured to improve products and identify needs.

### Comments from CEO Masanori Morishita

Call centers are the area where voice AI can provide social value fastest. However, one must be cautious of the 'everything is AI' argument. The essential issue over the next five years is drawing the line between tasks for AI and tasks for humans.

We aim to be infrastructure that enables people to focus on their core work rather than replacing them. Following this call center analysis, we will sequentially publish market forecasts for major industries like hotels, healthcare, and F&B to visualize 'substitutable' and 'non-substitutable' tasks.