Working every loan drop-off — and converting 35% above the human team

How a leading NBFC re-engaged the top-of-funnel leads its agents never had time to call.

Industry

NBFC / lending

Use cases

Loan sales, drop-off re-engagement

Languages

11+ Every state in India

Achievements

↑ 35%

higher conversion than human agents, on the same leads

↑ 40%

higher connection rate than human agents

100x

the lead volume covered — every drop-off, the whole base

Someone starts a loan application, gets to the income page, and closes the tab. By the time a human agent works down the list to that lead, the intent is gone — and most leads like it never get called at all; there was never enough time. The NBFC wanted to know what happened if every one of them got a call back, fast, in their own language.

Challenge

The leads that go cold before anyone calls

Volume beyond reachIntent has a shelf lifeThe conversation is the bottleneck

Lending’s top of the funnel is enormous and impatient. A team works the leads it can reach in a day, and the rest cool off — not lost to a competitor, just to a call that came too late or never came.

Reaching them wasn’t enough on its own. The agent had to answer real eligibility questions, hold a borrower through the awkward parts, and get them to a submitted application — without ever sounding like a script.

How we built it

A call back the moment the lead drops

The agent calls the second someone drops off, picks up where they left the form, and answers the eligibility and rate questions that stalled them — pulled live from the lender’s own loan documents. When the borrower is ready, the application or payment link goes out on the call.

If now isn’t the moment, it sets a callback for when the borrower said to try again. And it carries the whole conversation in the Hindi-English mix people actually use, so nothing stalls on language.

A real call

Pre-qualified loan sales
(Hindi)

App guided journey sales
(Hindi)

Results

It beat the team on the team’s own leads

Start with reach, because that’s the part a roster can’t fix: the agent worked every drop-off, the full base, where a team only ever got to a slice. On the leads both did call, it connected 40% more often. And the number the whole bet rested on — conversion — came back 35% higher than the human agents, on the same leads.

How we measured

Connection and conversion are measured against the human team on the same lead pool, same period. Connection rate = reached and engaged; conversion = completed application or captured intent.

What's next

Built to scale across India’s languages

The agent now operates in 11+ languages across every state, and does far more than translate. Customers switch between Hindi and English, ask half-finished questions, and change their minds — and the agent follows all of it the way a person would.

Arrowhead AI calling bots

Speaks like a human.Performs like a machine.