.svg){kind=small}
%20(1).webp)
AI Won't Fix Hotel Marketing Until Your Data Is Connected
Most hotel groups have now run their first AI experiment. A tool was connected to the ad accounts, produced confident-sounding output, and six weeks later nobody in the commercial meeting could say whether margin actually moved.
That is not a failure of the model. It is a failure of the data underneath it.
The evidence outside hospitality is blunt. McKinsey found that nearly two-thirds of enterprises worldwide have experimented with AI agents, yet fewer than one in ten have scaled them to deliver tangible value, and eight in ten name data limitations as the roadblock. Hotels are a textbook case.
The pattern is consistent enough to predict.
Two systems counted two different things, and the AI picked one of them. A digital hotel marketing programme and a revenue team can both be running well and still produce two irreconcilable versions of the same month.
The scale of the measurement problem is not a matter of opinion. Google research followed a single traveller who engaged with more than 500 separate digital touchpoints while researching one flight.
One person, one trip. Now consider what your systems retain of it: a last-click attribution, a booking record, and nothing in between.
Skift's 2026 Data and AI Summit research adds a gap at the top of the funnel. Eighty-two per cent of AI hotel recommendations come from OTAs and editorial media, while only 6% of hotels appear in AI results at all. Closing that gap, the research concluded, requires marketing, revenue, distribution and PR to operate as one integrated function.
What AI can and cannot see in your marketing data
It can see: impressions, clicks, cost, and the conversions the ad platform claims credit for.
It cannot see:
Digital hotel marketing produces enormous volumes of signal. Very little of it is joined to the outcome that pays the wage bill, which is a confirmed, arrived, paid-for stay.
The blind spot runs both ways. Your hotel revenue management process knows exactly which rooms sold at which rate, and nothing about which paid demand delivered them.
pulse. closes that loop. Ad performance is read alongside the booking record from the PMS, so the number you act on reflects the stay rather than the click.
Your commercial function is measured on four different numbers.
All four can hit the target in the same month while the group earns less per room-night than it did the year before.
This is why a strong hotel marketing strategy and a disciplined approach to hotel revenue management can coexist with falling profitability. They optimise different things, and no shared number tells them when they are working against each other.
The margin for error is narrowing. Deloitte's 2026 Travel Industry Outlook reports rising financial caution across all income levels, with 53 per cent of frequent corporate travellers planning three or more trips a month, down from 63 per cent the year before.
For owners, operational wins that never reach the asset valuation are the most expensive blind spot in the business.
Useful hotel data analytics starts with connection, not visualisation. Four tests to run on the next tool you are pitched.
Marketing reports by the date of the click. Revenue reports by the date of the stay. Until both sit on the same timeline, comparing them produces a number that looks precise and means very little.
In pulse.: all three data sets land on a shared timeline before any recommendation is generated.
An ad platform reports a conversion. The PMS reports a reservation, then a cancellation, then a no-show, then an arrival.
Serious hotel data analytics reconciles the first back to the last. Anything else is counting intentions.
In pulse.: attributed outcomes are reconciled against the PMS.
Some figures are measured. Some are calculated. Some are a snapshot. Some are an estimate. All four are legitimate. Presenting them identically is not.
In pulse.: every figure carries its origin, so you can see whether you are reading measurement or judgement.
Connecting the data still leaves three arguments in a nicer interface if marketing is scored on return, revenue on RevPAR and distribution on mix. Adding hotel revenue management software sharpens the pricing decision without settling that argument.
In pulse.: every commercial signal contributes to every commercial decision. Marketing informs pricing. Pricing informs marketing.
Before, the first twenty minutes go on reconciling whose figures are right, and decisions get deferred to next week.
After, with pulse. on screen:
That is the point at which a hotel marketing strategy becomes testable rather than defensible. Smaller leaks surface too, including cart abandonment inside the booking journey.
pulse. by dhi Hospitality connects marketing, revenue and property signals into one commercial layer, then turns them into the next move to make.
pulse. reads your PMS, your rate intelligence, your channel manager and, where you have one, your hotel revenue management software. Nothing gets ripped out. Setup takes about two weeks when your stack is already connected.
Skift's research made the same point from the operator side: deploying AI is the starting point, and value comes from governance, monitoring and human oversight on decisions that carry risk.
Apply the same standard to any hotel revenue management software or digital hotel marketing tool you are shown.
AI is not the constraint in hotel commercial performance. Disconnected data is.
Once your hotel marketing strategy, hotel revenue management and property signals sit on one timeline, reconcile to one confirmed booking and answer to one goal, hotel data analytics stops describing the past and starts shaping the next decision. Until then, a model repeats your existing disagreements back to you at speed.
pulse. launches on 30 September. Early access is open to hotel groups who want to move first.
.webp)
.webp)