The brief, and the wrong answer
A luxury hotel asked me to grow its direct bookings. Everyone assumed that meant a bigger budget.
Most of the property's business ran through the online travel agencies. That meant a commission on every room and almost no control over the guest. Direct bookings, the ones the hotel keeps in full, were flat.
When bookings are flat, the instinct is to spend more. But more budget on a broken funnel does not fix the funnel, it just drains faster. So before touching the budget I looked at where it was already going, and almost none of it was reaching the people who actually book a room like this.
Why hotel campaigns waste money
Most chase reach. They buy impressions from anyone who has ever looked at a beach photo, because reach is cheap and it feels like progress.
But a luxury room is not an impulse buy. The person who books it researches, compares and decides over days. Spread the budget thin across everyone, and the few people who were ready to book never see enough of you to trust you. You pay to be seen by people who will never book, and you underspend on the people who almost will.
What I changed, in order
Cut the broad awareness spend that was producing impressions and nothing else.
Rebuilt the targeting around the segments that had converted before: specific source markets, real intent signals, people already deep in the booking decision.
Shifted budget toward direct bookings, so every sale kept its full value instead of handing a cut to an agency.
Optimised every week rather than every month. Each week, money moved away from what was not converting and toward what was.
What came out of it
Revenue +423%
Conversions +348%
Transactions +113%
The two numbers I care about more than the headline are the second and third. Together they say the growth came from more people booking, not from a handful of large orders flattering the average.
The honest part
The 423% came off a low base. A jump that size is only possible when the starting point is small and the foundation is weak. A property already doing strong numbers will not repeat it. The method is the same at any size: find the budget being wasted, and move it to where it converts.