How Flexcar Scaled Phone Support 300% Without Hiring More Agents Using Voice AI
Most growing companies face the impossible choice between degrading customer experience or exploding support costs. Flexcar found a third way - automating 60% of calls with voice AI while improving satisfaction scores. Discover their phased implementation that maintained human touchpoints for complex issues while handling routine inquiries at scale.
The Support Scaling Challenge
Flexcar, a month-to-month car subscription service, faced a critical inflection point in . Their customer base had grown to 7,000 active subscribers across multiple cities, but their support team remained small - just 18-20 daytime agents handling all phone and email inquiries. With expansion plans targeting new markets, leadership knew something had to change.
"We were staring down 300% growth projections," explains Leslie Chen, Flexcar's Head of Customer Experience. "The math simply didn't work - either we'd need to triple our support staff, degrade service levels, or find a smarter way to scale."
Key Insight: 60% of Flexcar's support calls came through phone channels, yet their traditional IVR system couldn't intelligently route or resolve inquiries. This created bottlenecks during peak hours while overnight agents sat idle waiting for emergency calls.
Limitations of Traditional Phone Support
Flexcar's existing phone support system suffered from three critical limitations that made scaling impossible:
- Inflexible staffing: Their 8am-5pm support window missed 42% of customer inquiries that came after hours when people researched car options
- High variability: Occupancy rates swung wildly from 90%+ during day shifts to under 15% overnight
- Manual processes: Agents spent 30% of call time on repetitive tasks like looking up account details or reading scripted responses
"Our overnight team would wait hours between calls," Chen notes. "But we couldn't reduce staffing because those 3am emergency calls absolutely needed human response."
Why Voice AI Became the Solution
Flexcar identified voice AI as their scaling solution after analyzing call drivers. They discovered:
- 30% of calls were from new customers asking basic "how it works" questions
- 22% were routine billing or account status inquiries
- Only 15% required specialized human knowledge or emotional intelligence
"We realized most calls followed predictable patterns," Chen explains. "When an agent looked up a Guru article and read it verbatim, that was our signal the interaction could be automated."
Implementation Insight: Flexcar's phased approach started with AI handling just 10% of call volume, gradually expanding as confidence in the technology grew. This minimized risk while allowing continuous optimization.
Phased Implementation Strategy
Flexcar's three-year voice AI rollout followed this deliberate progression:
Year 1: Agent Assistance Tools
AI provided real-time knowledge base suggestions during calls but didn't interact directly with customers. This built internal confidence in the technology's accuracy.
Year 2: Silent Monitoring
AI listened alongside agents, automatically creating tickets and suggesting resolutions. This captured call driver data that informed future automation priorities.
Year 3: Full Call Handling
Voice AI took ownership of routine inquiries while seamlessly transferring complex cases. The system handled authentication, data lookup, and simple resolutions without human involvement.
"By year three, we'd built the knowledge base and confidence to let AI run independently," Chen notes. "But we maintained human override capabilities for any edge case."
Primary Voice AI Use Cases
Flexcar focused automation on these high-volume, low-complexity interactions first:
1. New Customer Onboarding (30% of calls)
AI answers FAQs about subscription terms, vehicle swapping, and pricing models
2. Billing Inquiries (22% of calls)
Handles payment questions, due dates, and receipt requests without agent involvement
3. Appointment Scheduling (18% of calls)
Books maintenance visits by checking vehicle availability and customer calendars
4. Account Status Checks (10% of calls)
Provides subscription details, reward balances, and contract terms on demand
Key Metrics and Success Indicators
Flexcar tracked these KPIs to measure voice AI effectiveness:
| Metric | Before AI | After AI |
|---|---|---|
| Average Handle Time | 8.7 minutes | 2.3 minutes |
| Containment Rate | 0% | 60% |
| Customer Satisfaction | 82% | 94% |
| Overnight Staffing | 4 agents | 1 agent |
"The biggest surprise was improved CSAT," Chen reveals. "Customers appreciated 24/7 access for simple questions rather than waiting for business hours."
Impact on Support Team Structure
Voice AI transformed Flexcar's support organization:
- Role specialization: Daytime agents shifted from generalists to specialized roles handling complex cases
- Overnight reduction: Cut overnight staff by 75% while maintaining emergency response capabilities
- Quality focus: 40% more time spent on high-value interactions rather than routine queries
"Our best agents actually preferred this model," Chen notes. "They hated repeating the same scripts and could focus on truly helping customers."
Future Voice AI Roadmap
Flexcar plans these voice AI expansions in :
- Image analysis: Processing accident photos to automate claims
- Omnichannel handoffs: Continuing conversations across phone, app, and email
- Predictive support: Proactively contacting customers about upcoming billing or maintenance needs
"We're just scratching the surface," Chen says. "Voice AI will eventually handle 80% of contacts while making the remaining 20% more meaningful."
Watch the Full Tutorial
See Flexcar's voice AI implementation in action (jump to 12:30 for the demo of their custom workflows).
Key Takeaways
Flexcar's voice AI journey offers three critical lessons for scaling support:
In summary: 1) Start small with predictable use cases, 2) Measure both efficiency and experience metrics, and 3) Continuously expand automation boundaries while preserving human oversight for complex scenarios.
Frequently Asked Questions
Common questions about voice AI for customer support
Flexcar automated 60% of their inbound call volume using voice AI, focusing primarily on routine inquiries like billing questions, appointment scheduling, and account status checks.
This allowed their human agents to focus on more complex customer needs while maintaining the same team size despite 300% growth in customers. The automation rate increased gradually from an initial 10% as confidence in the system grew.
- Highest automation rates: billing inquiries (85%) and appointment scheduling (78%)
- Lowest automation rates: accident claims (15%) and contract disputes (10%)
- Average handle time for automated calls: 2.3 minutes vs 8.7 minutes with humans
Despite initial concerns, Flexcar saw a 12% improvement in customer satisfaction scores after implementing voice AI.
Key factors included faster resolution times for simple inquiries and 24/7 availability for basic support needs. The AI system maintained consistent response quality regardless of time of day or call volume spikes.
- CSAT for automated calls: 94%
- CSAT for human-handled calls: 96%
- Biggest satisfaction driver: reduced wait times (cited by 68% of customers)
Flexcar prioritized three initial use cases that accounted for 60% of total call volume with predictable resolution paths.
They focused on inquiries where agents typically referenced knowledge base articles or followed scripted responses. This created quick wins while minimizing risk of poor customer experiences.
- Routine billing inquiries (30% of calls)
- Appointment scheduling for vehicle maintenance (20% of calls)
- Account status checks (10% of calls)
The voice AI system authenticated callers through API connections to Flexcar's customer database, verifying information like phone numbers and account details.
For security-sensitive actions, the system would transfer to human agents who followed existing authentication protocols. Flexcar maintained strict data privacy standards throughout the implementation.
- Basic authentication success rate: 92%
- Average authentication time: 23 seconds
- Failed authentications automatically routed to human agents
Flexcar tracked five key metrics to evaluate their voice AI implementation, focusing on both efficiency and customer experience.
These metrics helped them continuously optimize the system while ensuring quality standards were maintained as automation rates increased.
- Containment rate (percentage of calls fully resolved by AI)
- Transfer rate (calls needing human escalation)
- Average handle time
- Customer satisfaction scores
- First call resolution rate
Flexcar maintained their core support team but reduced overnight staffing by 75% as voice AI handled most non-emergency nighttime inquiries.
Daytime agents shifted focus to complex cases, with 40% more time spent on high-value customer interactions rather than routine queries. The team developed specialized roles for different types of escalations.
- Overnight team reduced from 4 agents to 1
- Daytime agents gained 2+ hours daily for complex cases
- New specialized roles created for technical and emotional support
Flexcar implemented voice AI in three phases over three years to ensure smooth adoption and continuous improvement.
This gradual approach allowed them to build internal confidence in the technology while collecting data to optimize the system before full automation.
- Year 1: Agent assistance tools (AI suggestions during calls)
- Year 2: Silent monitoring (AI listening alongside agents)
- Year 3: Full call handling for routine inquiries
GrowwStacks designs custom voice AI solutions tailored to your specific support needs. We analyze your call drivers, build targeted automation workflows, and integrate with your existing CRM and knowledge base systems.
Our implementations typically automate 40-70% of routine inquiries while maintaining seamless human escalation paths. We follow Flexcar's proven phased approach to ensure successful adoption.
- Free call driver analysis to identify automation opportunities
- Custom workflow design for your unique use cases
- Seamless integration with your existing support stack
Ready to Transform Your Phone Support with Voice AI?
Scaling support headcount linearly with customer growth is unsustainable. Let us help you implement Flexcar's proven voice AI strategy to handle 60%+ of calls automatically while improving customer satisfaction.