Realtor and co-broker yield model

Every co-broke dollar, bonus and referral joined to the agent who earned it and the closings that agent has actually delivered. The channel works. The schedule on top of it has never been written.

$1.10M

Reference-scale value

Data sets in the model

6 weeks

Build and deploy

Derived

Evidence grade

The co-broke is not the problem and we will not argue it down

A hundred and seventy-one closings a year arrive through an agent. At two and a half per cent that channel costs $2.90M and it is money well spent — a builder who trims the rate finds out inside two quarters what it costs in showings.

So the rate is out of scope, permanently. What is in scope is everything paid on top of it, which nobody has ever tested against yield.

A hundred and twenty-eight agents brought you a buyer in three years. A hundred and six of them brought exactly one and have not been back.

The bonus above standard co-broke is the same $2,999 whoever writes the closing. That is not a pricing error; it is a schedule nobody has had the data to write.

Cause 01

Bonus with no yield test

A flat bonus above standard co-broke paid on every closing. Forty per cent of it buys real placement and is not claimed.

Cause 02

Spend against a channel tail

Sixty-two per cent of co-broked closings come from agents with a single closing in three years.

Cause 03

Incentive stacked on a full co-broke

Fifty-eight closings taking both, at $6,800 each. Credited to the incentive model and not claimed here.

Cause 04

Commission paid twice

Eleven closings a year settled on procuring cause without anybody reading the registration log.

Want more detail?

The rest of this model — the data sets it runs on, the cause split, the working behind the reference figure and a sample output — goes out by email. Two fields, nothing else.