AI Lead Qualification Assistant for Housing Loan

Banks and NBFCs often face a significant bottleneck in their home loan sales funnel, with lead-to-conversion rates typically remaining in the low single digits. Some of the primary reasons being delayed outreach after lead allocation to a Feet-on-Street (FOS) agent, lack of lead context during customer outreach and missed engagement windows during the critical first few minutes/hours after a customer expresses interest. In many cases, FoS agents typically handle 10-15 inbound leads per day, making it difficult to consistently reach every lead quickly and conduct structured qualification conversations.

To address this challenge, we have designed a Voice-based AI Lead Qualification Assistant. The solution is an automated, multilingual voice agent that immediately engages newly assigned leads, conducts qualification conversations, scores lead quality and routes prospects to the appropriate downstream teams.

The workflow is as follows:

• A new lead is generated after a lead fills out the registration form

• Within seconds, the AI Voice Agent initiates an outbound call. The AI agent introduces itself as the bank's virtual assistant

• The AI Agent conducts a structured 3-5 minute qualification conversation in the lead's preferred language

• Information gathered during the conversation is evaluated against predefined qualification parameters, which are configurable. A score is given

• Based on the score, the lead is categorized as : Hot, Warm, Cold, Pending (conversation deferred), Urgent (immediate assistance required)

• A ticket is automatically created in the CRM containing the conversation summary, qualification details, lead score and routing recommendation

• Hot, Warm and Urgent leads are routed to FoS agents for further engagement. Cold and Pending leads are routed to the Quality Audit team for review and follow-up

The objective of the solution is to ensure immediate lead engagement, improve lead qualification consistency, increase FoS productivity and improve overall conversion outcomes by prioritizing high-quality prospects.

Right Time to Contact Leads

Since the 2007 MIT Lead Response Management Study of Dr. James Oldroyd, it has been known that the best time to make the first contact with a lead is within the first five minutes after the lead showing an interest in your product or service. That's a small window and once it's closed, the odds of contacting the lead decrease precipitously. The drop is so steep that the odds of contacting a lead decrease 100 times if you delay the contact by 30 minutes as opposed to five minutes. The delay also decreases the odds of qualifying the lead. Suppose a lead shows an interest in a home loan offer at 04:00 p.m. and you call them at 04:05 p.m., then you are 21 times more likely to qualify the lead than if you were to call them at 04:30 p.m.

Logistics of Calling at the Right Time

A Feet-on-Street (FoS) agent is expected to reach out to 10-15 inbound leads on each working day. However they cannot reach out to every lead assigned within a 5-minute window.

That's a human bottleneck. For this reason, we recommend a solution that automatically qualifies incoming leads and passes only the qualified prospects to the FoS agents, allowing them to focus their time and effort on high-value conversations and ultimately drive conversions.

The Voice-based AI Lead Qualification Assistant is an automated, multilingual voice agent that can engage newly assigned leads, conducts qualification conversations, scores the lead and routes prospects to the appropriate downstream teams; and it does all of it within five minutes of the lead showing an interest.

A lead shows an interest by filling a short form as shown in the next image.

As soon as the lead presses Submit, the AI Lead Qualification Assistant, using a webhook, fetches the lead information from your CRM and places a voice call to the lead. During the call, it enquires about the lead's loan requirements, property details, lead's employment details, and it possesses the ability to respond to any queries the lead may have. Once the information exchange is complete, it confirms the date and time when the FoS agent should reach out to the lead and end the call.

Categorizing Leads as Hot, Warm, and Cold

Based on the conversation, the AI Lead Qualification assistant will categorize a lead as Hot, Warm, or Cold. This categorization helps the FoS Agent prioritize leads based on their business value.

The Hot and Warm leads are transferred to the FoS Agent. The Cold leads are sent to the Quality audit teams for further qualification.

A lead can be categorized as Pending if the agent doesn’t have sufficient information from the customer to score them, including a call drop. Certain leads can also be categorized as Urgent in case they show some form of urgency during the conversation, such calls will be immediately routed to the FoS Agent.

The categorization is based on five parameters, which are listed in the next table. Please note that the parameters, weightages and the scoring criteria are configurable.

Parameter	Weightage	Scoring Logic (4→1)	Theme of Questions
P1: Purchase Intent & Timeline	30%	4=Ready (<3 months) 3=Shortlisted (3-6 mo) 2=Exploring (6-12 mo) 1=Browsing	The agent asks questions about the status of property and when the lead intends to buy.
P2: Loan Eligibility Signals	25%	4=Salaried stable, low liabilities 3=Self-employed with proof 2=Irregular/high liabilities (>50%) 1=Unclear/refuses	The agent asks questions about the lead's employment status, income, current EMIs, and PAN details.
P3: Loan Amount & Property Value	20%	4=₹30L- ₹2Cr 3=Ticket outside range 2=Vague 1=No clarity	The agent asks questions about the property's price and the lead's loan requirements.
P4: Decision-Making Authority	15%	4=Primary Decision Maker 3=Joint with spouse (common in India) 2=Family approval needed 1=Not the Decision Maker	The agent asks questions about who is going to make the final call.
P5: Preference	10%	4=your company is the first preference 3=Comparing 2-3 incl. your company 2=Primarily PSU/NBFC 1=Already finalized elsewhere	The agent asks questions about the lead's preferred lenders.

Lead categorization status and thresholds

Category	Score Range	Routing	SLA for human agent (indicative, can be configured)
HOT	80-100	FoS Agent - Immediate	Callback within 5-10 minutes
WARM	55-79	FoS Agent - Scheduled	Callback within 2 hours
COLD	0-54	Quality Audit Team	Nurture drip / re-qualify in 30 days

First Contact from AI Lead Qualification Assistant

This section illustrates the first contact from the AI Lead Qualification Assistant to a lead. The agent workflow during the call is illustrated in the next image:

Introduction

The agent greets the lead and confirms their identity at this stage. In this illustration, the agent calls the leads and speaks in Hindi. Support for more Indian languages are available.

Purchase Intent and Timeline

The AI agent asks questions about the status of the property (whether it is shortlisted or they are still searching) and the timeline of the loan. Based on the inputs, the AI agent tries to understand the urgency of the applicant. Follow-up questions on RERA registration are also asked which indicates the lead’s urgency to get a home loan.

Feature : The lead can digress and ask an off-flow question either related to the home loan, or any other offerings. The AI Lead Qualification Assistant is also trained on knowledge base and responds only from the knowledge base. After the response is given, it swiftly brings back the lead back to the qualification flow.. For example, if the lead asks about the tax benefits on a home loan, the agent finds an answer from the company's knowledge base and responds with the same it to the lead.

Loan Eligibility Signals

At this stage, the agent asks a few questions about the lead's employment history, monthly take home income, and monthly obligations. The agent also asks for the lead's PAN number to check the lead's credit profile.

Loan Amount & Property Value

The AI agent asks a few basic questions about the loan amount required and the value of property they are planning to buy. If these are mentioned explicitly during the conversation already, these questions will be skipped.

Decision-Making Authority

In this stage, the AI agent will check if the lead wants to apply with a co-applicant which further strengthens the application. It will also swiftly nudge the customer to understand whether they are the decision-maker or not.

Preference & Competition

The AI agent enquires if the lead is looking for a loan from their company, or if the lead is also looking for competing offers from other banks and NBFCs.

Soft Sell & Close

In this stage, The AI agent does a soft sell and tells about the benefits of taking a home loan from the bank/ NBFC. It also suggests a few home loan offers and gives an overview of the process ahead.

During this stage, the AI agent also confirms the time when the next call should be given to the lead for further processing of the application.

Once the call is over, the AI agent creates a ticket in your company's CRM. Based on the lead's score on five parameters, the agent assigns the ticket either to an FoS agent or the Quality Team. The ticket contains all the information an FoS agent or the quality team will need to proceed with the customer.