$0 until it works. You pay only for booked meetings and reactivated customers — never for calls or minutes. See how →
PTS AI › Blog

Branch-Level Reporting for Lenders: What a COO Should See Weekly

The weekly branch board a lending COO needs to see if calling works: the chain from call to funded loan, the ratio that matters, and one ID tying it together.

By the PTS AI teamSeptember 18, 20266 min read
The short answer

Every week, a lending COO should see each branch on its own row: calls placed, customers reached, interested, transferred live, visited the branch, and loans funded. Judge calling by live transfers per 1,000 calls, not dials. And tie every step to one ID from the first call to the funded loan, so you can prove what produced the revenue.

Every week, you should see each branch on its own row. Not one total.

On that row: how many people were called, how many were reached, how many were interested, how many were transferred live to the branch, how many walked in, and how many loans funded.

Judge the calling by one honest ratio: live transfers per 1,000 calls. And make sure one ID follows each customer from the first call to the funded loan. Without that, you can guess what calling produced, but you can't prove it.

That's the whole answer. The rest of this is why, and what it looks like on a Monday morning.

The total looks fine. One branch doesn't.

You know how this goes.

A regional manager sends the weekly numbers. Calls are up. Transfers are up. Everyone nods.

Then someone asks about one particular branch. And it turns out that branch had almost nothing all week.

Maybe the phone line was forwarding to a voicemail nobody checks. Maybe the branch manager was out and nobody picked up transfers. Maybe the list for that branch was mostly old numbers.

The total didn't show any of it. The strong branches covered for the weak one.

This is the real problem with a single number. It tells you the network is busy. It doesn't tell you where to look.

And the branch that's falling behind is the one you needed to see first.

Why "dials" is a vanity number

Most calling reports lead with dials. Big number, nice chart, goes up and to the right.

But dials measure effort, not results.

A system can place thousands of calls to numbers that don't answer, numbers that are disconnected, or people who hang up in three seconds. Every one of those counts as a dial.

You can't take "dials" to your board. You can't take it to a branch manager. It doesn't say anything about whether a borrower ended up across the desk from a loan officer.

Here's the ratio we'd put at the top instead: live transfers per 1,000 calls.

A live transfer means the AI reached a real person, that person was interested, and the call was handed to someone at the branch while the customer was still on the line.

In the Sunset Finance pilot, that ratio was about 8 live transfers for every 1,000 calls placed. The pilot made 13,159 calls in its first 14 days and produced about 100 live transfers across 10 branches.

You're probably wondering why that ratio matters more than the raw count.

Because it's the only number you can apply to your own list.

If you know roughly how many calls it takes to put one interested borrower on a branch phone, you can look at your list and do the math yourself. You don't need to trust a vendor's projection. You can check it.

Your ratio won't be the same as someone else's. Your list, your offer and your markets are different. That's exactly why you want to measure it per branch, week by week.

The chain from call to funded loan

Here's the chain that matters. Six steps.

  1. Called. The AI placed the call.
  2. Reached. A real person picked up and talked.
  3. Interested. They wanted to hear more about the offer.
  4. Transferred live. The call was handed to the branch while they were still on the line.
  5. Visited the branch. They came in.
  6. Loan funded. The loan closed and the money went out.

Each step drops some people. That's normal.

What you're looking for is the step where one branch drops a lot more than the others.

And here's the useful part: each step points to something different to fix.

  • Few reached? Look at the list and the caller ID. Old numbers, or a number customers don't recognize.
  • Reached, but few interested? Look at the offer and the script.
  • Interested, but few transferred? Look at the branch phone. Is someone there to pick up while the branch is open?
  • Transferred, but few visits? Look at what the loan officer says on that first call. Did they set a time?
  • Visits, but few funded? That's your lending process, and you know it better than anyone.

Which numbers a branch can act on this week

Not every number is a branch's job.

A branch manager can't fix the list. They can't change the offer. Those belong to you or your operations team.

But they can do something about three things this week:

  • Picking up transfers. If transfers are ringing out, that's a staffing question for Tuesday afternoon, not a strategy question.
  • Turning transfers into visits. Did the loan officer book a time while the customer was on the line?
  • Turning visits into funded loans. That's their daily work already.

So when you share the weekly board, tell each branch which columns are theirs. The rest are yours.

Nothing kills a report faster than asking a branch manager to fix a number they don't control.

One ID from the first call to the funded loan

This is where most reporting quietly breaks.

The calling system knows who it called. The phone system knows who got transferred. Your loan system knows who funded.

Three systems. Three ways to spell a name. Two phone numbers for the same person.

If nothing ties them together, you end up matching spreadsheets by hand. Or worse, you stop at "transferred" and never learn whether any of it turned into a loan.

The fix is simple to say. One ID for each customer, created at the first call and carried all the way to the funded loan.

Then, when a loan funds, you can look back and see: this customer was called on this day, from this branch's number, reached, transferred live, came in, and funded.

That's attribution. It's the difference between "we think calling helped" and "here are the loans that started with a call."

If you want to go deeper on measuring this, the Academy module on proof is the place to start.

What a simple weekly branch board looks like

Keep it to one screen. One row per branch. Same columns every week.

BranchCalls placedReachedInterestedLive transfersTransfers per 1,000 callsVisited branchLoans fundedMissed transfersNote
Branch A—————————
Branch B—————————
Branch C—————————
Network total—————————

A few things about this board.

The total goes at the bottom, not the top. You read the branches first.

"Missed transfers" gets its own column. That's an interested customer who was handed to the branch and nobody picked up. It's the most fixable number on the page.

Sort by transfers per 1,000 calls, lowest first. The branch that needs you is at the top of the list, where you'll see it.

The "Note" column is for one line from the branch. "Manager on vacation." "Phone line down Wednesday." Context in plain words beats another chart.

Compare each branch to itself, week over week. A branch in a small town and a branch downtown won't match. Whether each one is getting better or worse tells you more.

At PTS AI, every call gets written down, outcomes land in the CRM the same day, and each branch gets its own report instead of one total nobody can act on.

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

Questions lenders ask

Why not just track total dials?

Dials measure effort, not results. A branch can place thousands of calls and put nobody in front of a loan officer. Live transfers per 1,000 calls tells you how much of that effort turned into a real conversation.

What ratio did the Sunset Finance pilot see?

About 8 live transfers for every 1,000 calls placed during the pilot, across 10 branches.

Why does one ID from call to funded loan matter?

Without one ID that follows the customer from the first call to the loan, you can't connect a funded loan back to the call that started it. Then you can't prove what calling produced.

Should every branch be judged on the same number?

Look at each branch on its own row. A good total can hide one branch that is falling behind, and that branch is the one you need to see.

The proof

A Georgia lender already did this. Here are the numbers.

43,000past customers sitting on the list
13,159calls in the first 14 days
224live transfers to branches, and counting
10 → 28branches, pilot to today

In the pilot, about 8 live transfers for every 1,000 calls — the one ratio you can hold up against your own list. It's what happened at Sunset Finance, not a promise. Read the case study · Run your own list through it →

The offer

One branch. 14 days. Every customer called.

  1. Week 1Your list cleaned. Local number, caller ID and agent ready for that branch.
  2. Week 2The branch starts calling. Interested borrowers ring that branch's phone, live.
  3. Day 14You have the numbers — called, reached, transferred — by branch. You pay only for results.

$0 until it works. Then you pay per result: a booked meeting or a reactivated customer. No install fee, no monthly fee, nothing for calls or minutes.

Your next step

Talk to us before the list gets any older.

✓No risk to your network: nothing up front, and you only pay for results — a booked meeting or a reactivated customer. No results, nothing to pay.

The branch that's behind this week looks fine inside the monthly total. By the time the month closes, the customers it didn't call have moved on.

Leave your number. Our AI calls you within minutes — you'll hear exactly what your borrowers would hear — and books a time with our team.

Someone is going to call your past customers. It should be the lender they already trust.

Sources
  1. Sunset Finance Case Study — 43,000 Past Customers, Reactivated — PTS AI
Keep reading