The problem
The target was lending more money profitably. Not clicks, not applications. Loans written, at economics the business could sustain, within the limits set by risk and the capacity of the operations team.
That changes what a good result looks like. An application that risk declines costs money and produces nothing. A spike that operations can’t process becomes a customer problem. Growth only counts if the rest of the business can carry it.
The thinking
Start from the target and work backwards. How many loans does the quarter need, what can each one cost, and which searches produce applications that get approved?
At this spend, small improvements compound. A better approval rate on one segment is worth more than a large lift in volume. The question was rarely how to get more. It was where the next dollar was worth the most.
The other half was speed. Every competitor had the same platforms. The advantage was how quickly we could see an opportunity, decide, ship the change and measure it.
The work
I owned the digital decisions: where budget went, what got built, what got stopped, and the case for more investment when the modelling said the return was there.
Financial models of the acquisition economics, so bids and budgets answered to what a loan was worth, not what a click cost. Google Ads tuned towards approval, not volume. Risk and operations in the plan from the start, with volume paced to what the business could approve and process. Internal tools built to shorten the loop from data to decision to shipped change.
The result
The commercial target was beaten every quarter.
The more durable result was the operating model. Analysis, decision and change moved fast enough to become an advantage in a market where everyone was bidding on the same searches.