AI Voice Agent for Online Events: Lunaga, 10,133 Calls
AI voice for event confirmation: 10,133 calls in 18 days, 770 confirmations, and a 47.46% answer rate. How to reduce no-shows without team involvement?
Discover how an online training school managed 10,133 pre-event confirmation calls with an AI voice agent: no team involvement, 16% conversion on received responses, and 770 confirmed attendees. ## TL;DR - 10,133 calls in 18 days with an AI voice agent for online events; zero human operators - 47.46% answer rate; above average for B2C outbound campaigns - 770 confirmed attendances; 16.01% conversion on responses (7.60% on total calls) - Outbound calling automation integrated with lead form, calendar, and tracking - Reduced workshop no-shows with an AI voice agent reminding registrants  ## The Context: Lunaga and the Challenge of “I signed up” → “I’m online” ### Who is Lunaga Lunaga is a UK-based company specializing in online language training. Their customer acquisition model revolves around three-day live workshops on Zoom, serving as the entry point to their commercial journey. Every connected participant represents a tangible opportunity; every empty seat impacts revenue. ### The Challenges of the Education Market (2025–2026) In a market saturated with free offerings, the average show-up rate for live events fluctuates between 20% and 40%. The issue wasn't the product itself, but the communicative “gap” between registration and the start of the event: logistical doubts, waning interest, and inconsistent reminders. ### Before AI: Pre-Event Bottlenecks Automated email and WhatsApp messages only partially addressed the need. Manual phone calls were inconsistent, not scalable, and difficult to measure. Three critical points: uncertain confirmation, logistical friction (Zoom links, times, access), and the impossibility of managing thousands of 1:1 interactions without draining resources. ## Implementing the AI Voice Agent for Online Events ### Phase 1 — Use Case and KPIs - Mapping touchpoints (email, WhatsApp) and confirmation gaps - Primary objective: explicit confirmation (goal_met=true) - Providing logistical information and gathering technical objections ### Phase 2 — Conversational Scripting - Focus on date/time, familiarity with Zoom, frequent objections - Warm, informative, non-salesy tone - Approach as an AI voice agent reminding registrants, not a sales call center ### Phase 3 — Technical Integrations - Automatic lead ingestion from the registration form - Sync with event calendar to personalize date/time during calls - Support channel for technical issues; Google Sheet for real-time tracking ### Phase 4 — Testing and Go-live - Sample testing: script, time slots, call attempt cadences - Optimization based on initial feedback - Scaling: 10,133 calls in 18 days; zero operators involved  ## Challenges Overcome: Automation that “Doesn’t Feel Like Automation” The positioning of the call was key: pre-event assistance service, not sales pressure. Result: an answer rate of **47.46%**, well above the B2C outbound benchmark (20–40%), and fewer rejections. The outbound calling automation worked because it was perceived as tangible value to ensure attendees didn't miss the appointment. ## Results (March 09–26, 2026) ### Key Performance Indicators | KPI | Value | |---|---| | Total Calls | 10,133 | | Answers (duration > 0 sec) | 4,809 | | Answer Rate | **47.46%** | | Confirmed Attendances (goal_met) | **770** | | Total Minutes Handled | 5,059 min | | Average Duration per Answer | ~6.56 min | | Conversion on Answers | **16.01%** | | Conversion on Total Calls | 7.60% | ### Operational and Commercial Impact - Weeks of human work absorbed by the agent: over **5,000 minutes** managed - Consistent and repeatable message across >10,000 contacts - Zero variable cost proportional to volume; immediate scalability  ## Qualitative Benefits: Beyond the Numbers - Consistent information for everyone: fewer misunderstandings and support tickets - Prevention of technical frictions (links, times, access) - Efficiency: ~**6.6 minutes** for each confirmed attendance ## Best Practices for Training Events ### Why AI Voice Works Pre-event activities require precision, consistency, and timing: a repeatable message, high volumes, and a binary objective (confirmed/not confirmed). With an AI voice agent, workshop no-show reduction becomes a measurable metric, optimizable with each iteration. ### Mistakes to Avoid - Positioning the call as sales rather than a service - Ignoring technical frictions (Zoom, links, times) - Failing to track goals (goal_met, completion rate) and cohorts for optimization ## Outlook 2025–2026: From Automation to Predictive Connecting the entire data chain — lead source → confirmation → attendance → post-event outcome — enables optimizations in timing, segments, and content. Those who don't adopt outbound calling automation remain limited in volume, consistency, and measurability; those who use an AI voice agent scale every edition without increasing costs.  ## Why DeepAgent DeepAgent's voice agents are designed for specific use cases like pre-webinar reminders. They offer rapid integrations, native KPI measurement, and consistent conversational quality. In Lunaga's case, the agent handled 10,133 calls; the team focused on the experience of the **770** confirmed participants. ## AI Doesn't Replace Humans, It Empowers Them Lunaga transformed its pre-event process into a scalable and measurable one: more attendees, fewer no-shows, zero manual effort. Want to replicate this in your funnel? Request a DeepAgent demo: leave your contact information and test the agent in real time. You will be called back from +39 068 345 191. ## Frequently Asked Questions ### How does an AI voice agent confirm participation? An AI voice agent makes automated calls, verifies the date and time, clarifies logistical aspects (Zoom links, access), handles objections, and records the status (goal_met=true/false). With outbound calling automation and calendar integration, it personalizes the message and sends follow-ups if necessary. Everything is tracked for subsequent analysis and optimization. ### What is its impact on reducing no-shows in online workshops? In the Lunaga case, the agent achieved a 47.46% answer rate and 770 confirmations, representing 16.01% of responses. Generally, workshop no-show reduction depends on lead quality, timing, and scripting. By standardizing reminders and overcoming technical frictions, it's realistic to surpass B2C outbound benchmarks. ### What integrations are needed to activate outbound calling automation? All you need are: a lead source (form or CRM), an event calendar for date/time, a support channel for technical escalations, and a spreadsheet/BI tool for tracking. The AI voice agent reminding registrants synchronizes data, personalizes calls, and updates outcomes (answers, goal_met, notes) for the team in real time. ### How can calls avoid being perceived as aggressive sales? Position the call as a pre-event assistance service: confirm attendance, clarify doubts, reduce access errors. Use an informative, non-persuasive script, and conclude with a logistical summary. The AI voice agent reminding registrants increases response rates when the value is clear and the goal is to facilitate participation, not to sell. ### What is the typical ROI of an AI voice agent for events? The ROI derives from more live attendees (more commercial opportunities), time saved (automated call minutes), and lower variable costs. With 10,133 calls managed and over 5,000 minutes automated, the impact on cost per attendance and LTV improves. Monitoring post-event conversions completes the economic picture.