The presentation lasted forty-five minutes. The math was clear. The mechanism was explained. The CMO understood both.
She went back to her desk and did nothing.
Not because she disagreed. Not because the numbers were wrong. Because the thing being described was going to take eighteen months to compound, and the current quarter was already complicated, and the ownership group had questions about the renovation budget, and eighteen months felt like a different administration’s problem.
The OTA commissions were manageable. The occupancy was holding. The RevPAR comp was acceptable. Nothing was on fire.
The Number That Changed the Conversation
Twelve months later the revenue manager flagged it in a Tuesday morning meeting.
He had been watching it for two quarters and had finally put it in a slide: OTA-sourced bookings as a percentage of total revenue, trending up. Not because the property was performing worse. Because the direct channel was flat and the OTA channel was filling the gap.
The CMO recognized the slide immediately. Same mechanism, different source. The presentation she had ignored twelve months earlier, now arriving through the revenue manager’s yield report.
The difference was that this time the number belonged to them. Not an agency’s projection. Not a framework applied to a hypothetical property. Their commission line, growing, without anyone deciding it should.
She said she wanted to revisit the acquisition conversation. The revenue manager said he remembered the presentation. His tone suggested he had not filed it under solutions.
She called a meeting anyway.
The Conversation Nobody Writes About
It took three meetings, not one.
The first was with the CFO. He came prepared with the right question and did not like the answer.
Eighteen months before meaningful compounding. Twelve months before attributable data on new guests acquired through an audience they did not own. Six months before the first campaign results would be interpretable enough to defend.
“So we’re spending money we won’t be able to evaluate for a year.”
“We’re spending money we should have spent two years ago. But yes.”
He did not approve it that afternoon. He asked what channel was getting cut to fund it. The CMO had anticipated the question. She had not anticipated how much ground she would have to give to answer it.
The SEM budget. The channel running a 4:1 ROAS on branded search: guests already looking for the property, already past consideration, already arriving. She was proposing to reduce spend on demand that was finding them and redirect it toward demand that did not exist yet in any channel they controlled.
The CFO stared at the trade for a long moment.
“If they’re already searching for us, why are we paying to be found?”
That question occupied another full meeting. The CFO eventually approved a partial budget: four campaigns over eight months, contingent on a formal review at month six. Not a commitment. A test with a defined exit. The branded SEM reduction went through. Two weeks later the property’s branded search impression share dropped eight points. The OTA channel appeared to absorb some of that. The revenue manager noted it without commentary in the following week’s report.
The second meeting was the harder one.
The revenue manager’s objections were not philosophical. They were operational and specific, and he had prepared for the meeting in a way that made clear he intended to win it.
He wanted to know whether the new audience would arrive in peak periods or shoulder season. Whether there was a rate floor in the targeting. Whether the creative would conflict with existing yield windows. And whether there was any way to verify that campaign-attributed bookings were not simply claiming credit for guests who would have booked through another channel regardless.
That last question, incrementality, took forty minutes.
He wanted a holdout group. A portion of the target audience receiving no outreach, against which the campaign results would be compared. Without that control, he said, they were not measuring performance. They were measuring coincidence.
He was not wrong. The holdout requirement added complexity the program had not been scoped to include. Implementing it pushed the first deployment by six weeks. The CMO absorbed that delay without comment.
The third meeting was with the marketing coordinator, who had been running email campaigns for three years and had built every one of them around people who had already raised their hand. Past guests. Inquiry contacts. Pre-arrival sequences. Post-stay follow-up. People who knew the property existed and had given some signal of interest.
The work she was now being asked to do was categorically different. The audience was people who had never heard of the property. The creative had to earn attention from someone with no existing reason to give it. The offer architecture that worked on a returning guest would not work here.
Her first subject line test failed. The second was better. The third was the one they used.
The First Campaign
The data integration was not clean.
The property’s reservation system and the acquisition audience file used different record formats. The marketing coordinator spent the better part of a week reconciling them. A booking category tagged one way in the engine appeared differently in the matchback layer. She caught it before the deployment. The delay cost another four days.
The campaign went out seven weeks behind the original schedule.
Six weeks after deployment, the matchback results came back. The revenue manager reviewed the holdout comparison before accepting anything.
The campaign group produced eleven bookings out of roughly twelve thousand reached. The holdout group, matched to the same size, produced three. The conversion rate in the campaign group was 0.09 percent against 0.03 percent in the control. The revenue manager called it directionally real and not yet definitive.
Total attributable revenue: $35,156. Average daily rate $940. Average stay 3.4 nights.
The CMO presented the numbers to the CFO. He asked for the cost per acquired booking. The answer was not spectacular on first-stay revenue alone. On contribution margin with a credible second-stay assumption it cleared the threshold the CFO had set. He accepted it without enthusiasm and reminded her that the month-six review was in the terms.
The CMO approved the next two campaigns.
The second campaign underperformed. The segment, a cohort with a summer travel pattern, was reached in late March, outside their planning window. Response was flat. Attribution was ambiguous. The holdout showed almost no gap.
The CFO raised the underperformance at the month-five check-in. He asked whether the SEM reduction had been worth it given the results so far. He was not hostile. He was doing his job. The CMO made the case to continue through month six before any decision. If the third campaign did not perform, she told him, she would recommend reverting the remaining budget to SEM. She meant it.
The third campaign performed closer to the first. Not the same result. Smaller lift, tighter margins. But enough to continue.
Eight Months Later
The direct channel percentage had moved three points over two quarters. The CFO acknowledged it at the month-six review and extended the test budget. He called it directional. He did not call it a success.
The revenue manager remained skeptical of the attribution methodology in the way that revenue managers remain skeptical of anything not directly visible in the reservation system. He continued questioning incrementality. He had also started demanding tighter controls over deployment timing so that campaign traffic would not, in his phrase, dump volume into weeks where he already had group commitments.
He wanted control over the sequencing. He did not trust that the marketing side understood the yield calendar well enough to deploy without causing rate problems. The property had not previously had a conversation about when to reach new domestic travelers relative to forward occupancy and rate strategy. That conversation was now happening.
Then the February pacing report arrived.
Canadian pace was soft. China had not recovered as the team had projected. European forward bookings for the summer were running fourteen percent below prior year. The revenue manager pulled the numbers and presented them without interpretation.
“What can we activate on the domestic side?”
Eighteen months earlier there was no answer to that question. The property’s domestic reach had been the house file: past guests, inquiries, a well-managed list of people who already knew they existed. No domestic audience they could deploy against independently when a specific corridor weakened.
Now there was one. Not large enough to replace the international shortfall. The revenue manager said that clearly. What he also said was that he could sequence three domestic deployments against the shoulder-period gaps before summer and that the cost structure was known.
It was not rescue. It was having a domestic audience they could reach when the forward report turned against them.
What the Commission Statement Shows
The OTA commissions are still there.
The line is smaller than it was trending toward two years ago. Three points of channel shift does not end a dependency. It reduces the rate of growth of one.
The CMO did not make a strategic breakthrough. She made a decision she had already understood was correct and then waited twelve months too long to make it, then spent eight months building something whose full value she will not be able to point to cleanly for another year.
The morning after the February meeting she approved the next deployment. The revenue manager had already sent the segment brief. He had built in a holdout group. He had flagged two weeks he did not want touched due to a group block.
The program is not finished. It is running.
