Automated Qualification of Existing Customers
The Problem
There is a massive pool of more than 4 million existing customers whose quality has become completely unclear due to decades of acquisitions. A significant portion of the records is outdated, old policies are inactive, or phone numbers are no longer assigned. If this unqualified data base is processed manually by agents, it results in massive scatter losses and legal risks. Operational queues become clogged with failed attempts, driving Dials per Hour upward while minimizing the Right Party Connect (RPC) Rate. The root cause lies in the lack of pre-segmentation within the legacy systems.
The Solution
An AI-powered voicebot takes over the automated outbound contact and prequalification of the insurance portfolio. Through a structured dialogue with three short survey points, the customer is approached and qualified regarding their current life situation. Processing is event-based: the voicebot not only records responses, but also deterministically classifies unreachable records, invalid phone numbers (bounce rate), and toxic legacy cases. The result is a fully cleansed, legally compliant database that enables measurably high reactivation rates for profitable products.
Interview: Qualification of Insurance Existing Customers
A conversation between an Operations Manager (J) and Dialfire (SA).
J (Energy Supplier):
“Our core problem was the black box in the CRM. We had four million dormant customers from 30 years of acquisitions, but we had no idea who was still active. When we assigned our agents to manually process old moped insurance policies, it was a complete waste of resources. The contact rate was in the single-digit range, and we constantly risked calling customers without valid opt-ins.”
SA (Dialfire):
“This is a classic inefficiency problem with unstructured insurance portfolios. CRM systems store the status quo, but they cannot verify it automatically at scale. This is where we use the AI voicebot as an upstream filter. Instead of wasting expensive advisor time sorting dead numbers, the system hands the data over to our taskflow engine for automated outbound processing.”
J (Energy Supplier):
“The manual process was simply no longer scalable for us. That’s why we implemented the survey through the voicebot. Three precise questions to determine the current life situation and potential needs. The challenge before was that agents had to painstakingly work through these simple questions with thousands of dead records.”
SA (Dialfire):
“This is exactly where automated qualification shows its strengths. The voicebot systematically processes the data inventory in a horizontally scalable manner. If the customer answers the three questions, we translate the intent signals — such as interest in disability insurance — directly into operationally usable metrics.”
J (Energy Supplier):
“That was the biggest gain for us in terms of data logistics. We didn’t just receive the qualified responses. The bot also automatically identified invalid phone numbers, permanently unreachable contacts, and legally sensitive cases without marketing consent. Our invalidity rate was extremely high at the beginning, but that was exactly what made the cleansing rate per shift possible in the first place.”
SA (Dialfire):
“This happens deterministically within the taskflow through the assignment of metrics such as status and status details. An invalid number immediately receives the corresponding final status, making your actual bounce rate transparent. A revoked opt-in is instantly moved into a dedicated campaign stage through the routing rule framework and is therefore isolated from further processing. These records are then returned cleanly tagged to your inventory system via webhook.”
J (Energy Supplier):
“The result is extremely valuable operationally. We are now operating on a completely clean existing customer database. Data quality is transparent, and we can direct our agents specifically toward segmented target groups. Since sales now only speaks with prequalified contacts for cross-selling, our effective conversion rate has increased massively.”
SA (Dialfire):
“That is the core function of the flexible processing layer. We remove latency and compliance risk from the process. Cleansing is performed machine-based and immediately creates structured, usable data for marketing — measurable through clear KPIs, fully traceable, and without having to rebuild your existing IT architecture.”