Live Transfer or Callback? What Lenders Should Do

Transfer them live. When a borrower on an AI call says "yes, I'm interested," their own local branch should pick up right then, while the borrower is still on the line.

Save the callback for when the branch is closed, or when the borrower asks for a different time.

That's the answer. Here's why it matters more than it looks, and how to tell whether it's working.

The "yes" doesn't last

Think about the last time you said yes to something on the phone.

You were in the mood. You had a minute. The thing made sense right then.

An hour later you were driving, or cooking, or dealing with a kid. If someone had called back at that point, you'd have let it ring.

That's your borrower. The "yes" is real, but it has a short shelf life. They're interested at 11:05 on a Tuesday because the call reached them at a good moment and the offer fit what they needed. That moment is the most valuable thing the AI call produces.

The research on sales leads points the same way. In "The Short Life of Online Sales Leads", published in Harvard Business Review in 2011, firms that tried to reach a lead within an hour were nearly seven times as likely to have a meaningful conversation as firms that tried an hour later.

That study was about web inquiries, not phone transfers. But the idea carries over. Interest fades by the hour. A callback starts the clock over on someone who was already warm.

What actually happens to a callback

Here's how a callback usually goes at a branch.

First, the AI call ends with "someone from your branch will call you." A task lands in the CRM.

Then, it's a busy afternoon. Two walk-ins, a payment dispute, a collections call. The task waits.

Then, someone calls at 4:30. Voicemail. They leave a message.

Then, the borrower calls back the next morning. The loan officer is with a customer. Now the borrower gets voicemail.

Finally, everyone gives up without ever deciding to. Nobody said no. It just stopped.

That's phone tag, and every lender knows it. There are three ways a callback leaks:

• The borrower cools off. Tomorrow they don't need it as much, or they found it somewhere else.
• The branch never calls. Not out of laziness. The task just sits under everything else that's louder.
• Nobody reaches anybody. Two busy people trying to find the same free minute.

Extra attempts don't fix it cleanly either. A 2007 report from researcher James Oldroyd and the sales software company InsideSales.com included a survey of 495 companies. It found that more call attempts that didn't reach the person went along with lower qualification rates. The same report's call data found that after about 20 hours, each additional dial actually hurt the odds of reaching someone. It was co-authored by a company that sold calling software, so take it as a direction, not a law. The direction is clear, though.

A live transfer skips all of it. The borrower is on the line. The branch picks up. The conversation happens now.

Why their own branch, and why only while it's open

Two rules make live transfers work. Most setups break one of them.

Rule one: send the call to the borrower's own local branch. Not a central queue. Not whoever is free three counties away.

That's the branch they'll walk into to sign. The staff there knows the town. They may already know the borrower. A first conversation with the people they'll actually see across the desk beats one with a stranger in a call center.

Rule two: only transfer while the branch is open and someone can answer. This sounds obvious. It's where most pilots fail.

If the AI calls at 7 p.m. and gets a "yes," there's nobody to hand it to. So the "live transfer" turns into a voicemail at the branch, which is a callback with extra steps.

The fix is simple. Only place these calls during branch hours. Then every "yes" has somewhere to go.

If you want to go deeper on where the machine should stop and the person should start, the Human handoff module in our Academy covers it.

When a callback is still the right move

Live transfer isn't a religion. There are times a callback is the right call.

• The branch is closed. A borrower who says yes right as the branch is locking up should get a scheduled callback for the next time it opens.
• The borrower asks for it. "Can you call me after 5? I'm at work." Then call after 5. Forcing a transfer on someone who can't talk just burns the "yes."
• The branch phone isn't answered. If a transfer rings and nobody picks up, the borrower shouldn't be left in the air. The call should fall back to a callback, logged, with a time.

The point is that the callback becomes the backup, not the plan. And when it happens, it needs to be tracked like a real appointment, with someone's name on it, so it doesn't sink under the afternoon.

The number to watch: live transfers per 1,000 calls

Here's the number we'd ask any lender to look at first.

In the Sunset Finance pilot, a network of consumer lending branches in Georgia, the voice agent Tommy called past customers about a pre-approved loan of up to $1,000. It placed 13,159 calls in the first 14 days. About 100 of those became live transfers, across 10 branches.

That works out to about 8 live transfers for every 1,000 calls placed.

You might be wondering why we'd lead with a number that sounds that small. It's because it's the honest one.

"Calls placed" includes everything. The people who didn't answer. The voicemails. The disconnected numbers. The ones who said "not right now." Every real list has all of that.

It's tempting to report transfers per conversation instead, or per "qualified contact." That makes the percentage look bigger. But you can't plan with it, because you don't know in advance how many conversations your list will produce.

You do know roughly how many calls your list will take. So you can do the math yourself. Take the calls, divide by 1,000, multiply by 8, and you have a starting assumption for how many borrowers would land live on your branches' phones.

It also keeps everyone honest inside your own company. When a regional manager asks why one branch had more transfers than another, you can look at calls placed for each branch and compare like with like. A branch that got fewer calls should get fewer transfers. A branch that got plenty of calls and few transfers is telling you something, maybe about its hours, maybe about who's answering the phone.

A starting assumption, not a promise. Your list, your offer and your branch hours will move it up or down. That's exactly what a pilot is for.

How it runs at Sunset Finance today

Every transfer at Sunset Finance goes live to the customer's own local branch, while that branch is open. The call gets logged on its own, so the branch doesn't have to write anything down.

Since that first pilot, it's 224 live transfers and counting. The program has grown from 10 branches to 28.

A small way to test this

You don't have to change how your whole network works to see this.

Start with one branch. Within 14 days, every reachable customer on its list gets called, and the interested ones get transferred live to that branch. There's no upfront fee. You only pay for results: a booked meeting or a reactivated customer.

If you want to hear what your borrowers would hear, the AI will call you. Leave your number.