~35%

Reduction in Lead Qualification Costs

~30k

Automated Monthly Calls

In this case study

Fintech

Industry

Overview

Client overview

The client is a leading auto retailer with a strong online & offline presence. They also offer financing options to eligible customers.

For lead generation, they collect sign ups from multiple channels like Facebook, Google Ads, their own website,  etc. To qualify these leads, they make outbound phone calls to assess their interest and intent (eg: which vehicle they want to buy, by when, etc.)

But as the number of sign ups increased, the client was unable to scale the calling operations due to cost and operational constraints.

Challenge

The customer support challenge

The lead qualification process was largely manual. This led to inefficiencies at every stage from making initial calls, taking follow ups and connecting hot leads with sales agents. Hence, costs per qualified and converted leads were high. 

Some of the major issues they were facing are: 

High salary and training costs

  • Requirement to hire and train more agents as number of sign ups increased
  • Support for calls in multiple languages required hiring agents with specific language skills

No system to capture and store data in a structured way 

  • Calling data scattered across multiple agents (eg no. of calls made/ leads qualified)
  • No single dashboard to access a funnel view (channel wise bifurcation of no. of customers signed up, no. of customers called/ customers interested/ customers converted)

Losing qualified leads to competitors due to delays in taking next steps

  • Connecting hot leads with sales agents took up to 3-4 days 
  • 20-30% follow up calls were either delayed or missed due to bandwidth constraints

Solution

Robylon's solution

Here’s a step-by-step breakdown of how we developed and deployed our specialized AI voice agents:  

Step 1: Specialized AI agents to refine the calling SOP

A concise SOP was used as a starting point and given to the specialized SOP making AI agent. Call recordings of previous calls were also fed and a refined SOP was created. 

Step 2: Building, testing and deploying voice AI agents

We developed custom voice agents for the client to optimize their lead qualification process. The voice agents had the following features: 

  1. Human-like conversations
  • Ability to engage in 2-way conversations with natural speed and pauses while speaking
  • Custom tonality based on the purpose of calling (eg lead qualification v/s follow up)
  • Personalized responses based on the user’s questions (eg: describing the loan KYC process)
  1. Voice setup
  • Support for 15+ languages (English, Hindi, Kannada, Telugu, etc.)
  • 10+ variations in voices across male and female genders

Step 3: Setting up the calling operations

The following operational setup was made for the lead qualification: 

  1. Data fetching: Customer data (name, phone number, location, etc.) was fetched from the customer’s CRM via an API connector
  2. Outbound calling: Voice agents make calls to the leads and collect intent data (eg: vehicle model of interest, loan requirement) and also answer questions (eg: EMI amount, loan eligibility, etc.)
  3. Data capturing: The responses collected from the user were captured and updated real-time in the client’s CRM for data analysis and visibility (eg % lead conversion, etc.)
  4. Connecting hot leads: The AI agent can quickly connect the hot leads to a sales agent over the same call (eg when a customer requests or when a customer shows good interest)

Additionally, we also set up automation for making follow up and retry calls by AI voice agents. Follow up calls were made if a customer requested to call back at a certain time. Calls were also made after 15 days if a customer had not made a purchase by then. Lastly, up to 3 retry calls were made every alternate day if the calls remained unanswered. 

These targeted follow up and retry calls increased the lead qualification rates by 15%

Step 4: Single view dashboard

A single view dashboard with custom widgets was developed. It provided real-time access to business critical parameters like number of calls made/ leads qualified/ leads converted, channel wise analysis like conversion rates, etc.

Result

With Robylon’s AI voice agents, our client has been able to streamline the whole process of lead qualification. 

Prior to integration with Robylon, the client was able to make ~18k calls monthly. Over 4-5 months, Robylon helped scale this to ~30k monthly calls.

Due to lower costs of using AI voice agents over human agents, the lead qualification costs were reduced by over 35%

Additionally, the visibility over customer journey and the conversion funnel also provided valuable business insights like channel wise conversion rates, popular auto models, customer preferences, etc.

Other case studies