Outbound AI Voice Agent in Lending: Fincontinuo Case Study
Fincontinuo used an outbound AI voice agent to process 8,668 calls in 3 months, qualifying 560 lending leads and hitting a 37.5% success rate on live calls.
Borrowers requesting information on salary-backed loans expect an immediate response. In this case study, an outbound AI voice agent managed the initial outreach for consumer credit provider Fincontinuo. Over three months (May 1 to July 31), the agent executed 8,668 call attempts, completed 1,492 conversations, and handed off 560 pre-qualified leads to loan officers, achieving a 37.5% success rate on completed calls. ## Summary - 8,668 attempts over 3 months; 1,492 completed conversations; 560 pre-qualified leads. - 37.5% success rate on live conversations; 6.5% on total attempts. - Structured lead pre-qualification and automated scheduling keep loan officers focused on closing. - Direct outreach during evenings and weekends to reach prospects faster. - Call capacity scaled during campaign volume spikes without adding headcount. ## The client: Fincontinuo and salary-backed lending Fincontinuo is an Italian financial institution specializing in personal loans and salary-backed consumer credit deducted directly from payroll or pensions. Leads arrive through digital campaigns: users submit an online form and expect a rapid response. The target audience includes public and private sector employees as well as retirees. The goal was to reduce first-contact response times and provide credit advisors with verified applicant profiles rather than raw contact records. ## The challenge: reaching prospects without tying up advisors Three operational bottlenecks slowed the sales pipeline: - Reachability: Out of 8,668 call attempts, only 1,492 resulted in actual conversations. Voicemails and unanswered numbers accounted for most attempts. - Team workload: Loan officers spent hours on repetitive discovery questions instead of providing financial advice. - Response timing: Inquiries submitted outside standard business hours went cold quickly. In consumer lending, conversion probability drops as response time increases. ## The workflow: from web form to booked appointment Loan officers step in at the final stage, after the lead is pre-screened and a consultation slot is booked. 1) Campaign submission: The prospect submits an online form, and the lead enters DeepAgent into an automated outbound call queue. 2) Identity verification and availability: Giulia, the AI voice agent, verifies that she is speaking with the primary applicant. If the prospect is unavailable, she schedules a callback. 3) Financial pre-qualification: For employees, the agent records employer details, contract type, seniority, and net monthly salary. For retirees, she records pension category and monthly amount. In both scenarios, she verifies existing loan deductions and requested loan amounts. 4) Advisor handoff: The agent schedules a consultation date and time, confirms contact details, and sends a confirmation via SMS or email. The advisor receives a structured profile before the call. !Fincontinuo outbound AI voice agent figure 1 ## Funnel results: metrics and KPIs | Metric | Value | |---|---| | Call attempts (May 1 to July 31) | 8,668 | | Completed conversations | 1,492 (17.2%) | | Success rate on completed calls | 37.5% (1 qualified lead every 2.7 calls) | | Success rate on total attempts | 6.5% | | Total spoken conversation time | 52.5 hours | | Time per qualified lead | 5.6 spoken minutes | | Pre-qualified leads delivered to advisors | 560 | A positive outcome represents a pre-qualified lead or scheduled follow-up rather than a disbursed loan. The formal credit decision remains with the licensed loan officer. The AI voice agent qualifies prospects and fills the consultation calendar. !Fincontinuo outbound AI voice agent figure 2 ## Monthly trends: balancing volume and conversion Monthly data shows how volume shifts affect conversion rates: - May: 683 attempts, 60.2% success on completed conversations (103 outcomes out of 171). - June: 5,741 attempts, 33.1% success (279 outcomes out of 843), reflecting higher lead volume with wider variance in applicant readiness. - July: 2,244 attempts, 37.2% success (178 outcomes out of 478). Variables influencing outcomes included lead recency, campaign audience targeting, calling windows, and retry cadence. ## Conversational design: voice clarity and exception handling Prospects perceive the call as a direct response to their submission rather than a cold telemarketing dial. The voice agent references their submitted inquiry and adjusts pacing to the caller. Non-standard cases, such as active payroll deductions, contract renewals, or self-employed applicants, are not rejected outright. Instead, they are flagged and routed to advisors for secondary assessment. Once data collection finishes, the agent secures a calendar slot and sends an immediate SMS and email confirmation. !Fincontinuo outbound AI voice agent figure 3 ## Business impact on operations, costs, and borrower experience - Sales team: Advisors receive structured applicant summaries with income details, requested amounts, and notes so they can focus on underwriting and closing. - Operational costs: Call capacity handled 5,741 attempts in June without adding headcount or paying overtime. - Borrower experience: Applicants receive callbacks within minutes while purchase intent is high, securing consultations without waiting for the next business day. ## Key takeaways Across 8,668 attempts, the system produced 560 positive sales handoffs in three months, maintaining a 37.5% success rate on completed calls. An outbound AI voice agent handles initial outreach, verifies prospect qualifications, and delivers structured leads directly to loan officers. ## Frequently asked questions ### How does the AI agent integrate with existing CRM and calendar software? DeepAgent connects via API to standard CRM and calendar platforms. Data collected during the phone screening maps into structured fields, including verified phone numbers, salary details, and preferred contact windows. The consultation is booked directly on the advisor calendar with automated confirmation reminders. Webhooks update lead statuses and handoff outcomes in real time. ### How are compliance, data privacy, and call recordings handled? The setup complies with GDPR standards through caller notifications, data minimization, and configurable data retention policies. Audio recordings are optional and can be separated for quality assurance. Sensitive fields can be masked, and opt-outs or suppression lists are recorded automatically. Role-based access logs and Data Processing Agreements are available. ### What is the commercial model and expected ROI? Pricing is based on usage and capacity rather than per-seat licensing, establishing a clear cost per pre-qualified lead. In this study, 8,668 calls and 1,492 conversations yielded 560 pre-qualified prospects. By removing manual dialing and preliminary screening, sales teams reduce idle time, with return on investment determined by faster conversion and average deal size. ### How do you maintain conversation quality and success rates? The platform uses lending-specific conversation playbooks. The agent manages interruptions, confirms borrower identity, and validates required data fields. Real-time dashboards track connection rates, call durations, and completion metrics, allowing teams to adjust calling windows and opening scripts to maintain conversion rates. ### How does this change the daily routine for loan officers? Advisors spend less time manually dialing unverified numbers and more time in scheduled consultations. Each calendar invite contains income data, requested loan figures, and verified notes. Because initial outreach happens immediately, even after business hours, fewer prospects drop off before the initial consultation. Talk to sales → · Create your agent →