How a debt-recovery team automated 9 in 10 of its inbound calls

An Auckland debt-recovery firm turned its biggest daily distraction, the phone, into a system that mostly runs itself, and got about 20 hours a week back.

In debt recovery, the phone never stops. Jade Financial Services runs collections across New Zealand and Australia, and on a normal day their team takes 30 to 50 inbound calls. Almost all of them are the same handful of questions: what is my balance, when was my last payment, how much do I still owe, when is my next one due.

None of those calls need a skilled person. But every one of them pulled a skilled person off the actual recovery work to answer it. And to answer it, that person had to go digging through a spreadsheet to find the account first.

Why were 30 to 50 calls a day such a problem?

A debt-recovery team makes its money by working accounts, not by reading balances down the phone. Every routine call was an interruption: stop the recovery work, find the caller in the spreadsheet, read out a number, get back to what you were doing.

Across 30 to 50 calls a day, that added up to roughly 20 hours a week of skilled time spent on questions a system could answer. The data that should have made it easy, their customer database, made it harder, because it only existed as a spreadsheet nobody could query in real time.

What does the voice agent actually do on a call?

It answers the phone, verifies who is calling, and handles the routine account questions end to end, reading from live data.

It verifies before it shares anything

The agent identifies the caller one of two ways: their account number if they have it, or their name plus date of birth and address as verification. Only once the caller checks out does it pull up the account. For a finance business, that verification step is the security, built into every call.

It answers the questions that fill the day

Once verified, the caller can ask the things they always ask: how much is left on my account to pay, when was my last payment, what is my current balance, when is my next payment due. The agent reads the answer straight from their live account and gives it on the spot, any hour of the day.

It only escalates when it should

Around one call in ten is handed to a human. Anything sensitive, or outside the routine questions, gets escalated rather than forced through the agent. The team stops fielding the easy 90% and keeps the calls that genuinely need them.

Why move their data out of a spreadsheet?

A voice agent is only as good as the data behind it, and Jade's ran on a spreadsheet. Their entire customer database, across New Zealand and Australia, lived in Excel. Useful to a person clicking through it, useless to anything that needed to read it live.

We cloned that database into a secure cloud database with a daily sync, so it stays current without changing how the team works. Access is locked to owner and admin only. That one move did two things: it gave the voice agent live account data to read from, and it gave the business a real database to build on instead of a file that only worked one click at a time.

What was the real return?

The agent now handles roughly nine in ten of the 30 to 50 daily calls, which gives the team back about 20 hours a week, close to a thousand hours a year of skilled time returned to recovery work.

The work the agent does, fielding and verifying routine account calls, is what you would otherwise hire a receptionist for: around NZD $55,000 a year, working business hours only. The agent does roughly 90% of that volume, around the clock, for a fraction of the cost.

Why not just hire a receptionist?

Because the maths does not work. A receptionist to field those calls runs around NZD $55,000 a year all in, works nine to five, takes leave, handles one call at a time, and still has to look up every account in the spreadsheet by hand. The voice agent handles around nine in ten calls 24/7, verifies the caller and reads live data instantly, and never has a bad day on the phones. The team keeps the calls that actually need a person.

Could this be your team?

If your team is buried under inbound calls that mostly ask the same routine questions, or your business runs on a spreadsheet that can never become anything smarter, this is exactly the kind of problem we solve. We start with a Discovery Week: we find your single biggest win, scope it, and ship a first working piece fast, so you feel the value before any bigger decision.

Frequently asked questions

More case studies