01
The problem I had
At LendingTree the internet took the research and the mortgage application away from the sales rep. The brand promise was still on the air.
That brand promise is an important part of this narrative because it created the expectation that was never delivered once the customer met the salesperson.
The customer finished a long application and got one call with one rate. The ads had promised banks fighting for their business. So the customer went back out and shopped, and the company lost a sale it had already paid to attract. Nothing was wrong with the machine. The person's job had been left over, not rewritten.
02
What the pages show
Seven decisions, in the order that keeps the promise. The first two are the executive's own and cannot be handed off: what we promised, and where the customer's work now stops. The next five build the practice into the organization: the job on the far side of the machine, the order of spending, who does the two jobs a machine creates, the person's first ninety days, and the weekly meeting that keeps the line where it belongs.
Each page has the same parts. What you leave with. Three sections, three questions each, a box. My own answers as the worked example. Three steps, one action, and a prompt you paste into whatever AI you use so it interviews you and fills the page in.
03
What changed
At LendingTree the fix was not a new system. It was a decision about a role.
I made sure that they played the role of advisor. I would tell them to assume they are sitting on the same side of the table as the homeowner.
Write the person's job from the promise. Then build the machine to that job.