How Hotels Can Keep Price-Sensitive Bookings Direct

“I can see it cheaper on Booking.com.”

For a hotel, this is more than a price objection. It is the moment when a guest the property has already spent money and effort acquiring may leave the direct funnel — only for the hotel to pay an OTA commission if that same guest books elsewhere.

Marketing did its job. The guest found the property. Revenue management set the rates. The website, phone line or direct channel brought the conversation close to conversion. Then, at the final stretch, the guest hesitates over price.

If the direct booking journey has no useful response, the hotel risks paying someone else to close demand it already created.

That is the real business problem.

And it is also where you find out whether your booking technology can actually sell — or whether it is simply very good at displaying inventory.

The most expensive leak can happen one step before conversion

Most booking technology performs well when the guest follows the expected path.

“Do you have a room on Friday?”
“Yes.”
“How much?”
“€190.”
“Perfect.”

Availability retrieved. Rate quoted. Booking link sent. The workflow looks successful.

Real guests are less convenient. They hesitate. They compare. They ask whether the cancellation terms are the same. They remember a rate they saw on Booking.com. They decide that the room is right but the price is not. Or the price is right, but the conditions are not.

That moment is not a distraction from the booking process.

That moment is the booking process.

A guest who is comparing two offers is often closer to buying than a guest who has simply asked about availability. Yet this is exactly where many automated journeys become strangely passive.

The software knows the available room. It knows the rate. It may know the restrictions. But when the guest says, “I found it cheaper elsewhere,” the system often responds as if the conversation has reached an error screen: repeat the original offer, resend the link, or wait.

Technically, nothing has failed.

Commercially, the most important part of the conversation has just started.

What would an experienced reservations agent do here?

Probably not send the same booking link again.

A good reservations agent would first try to understand what the guest is actually comparing. Is it the same room category? Is one rate refundable and the other non-refundable? Does one include breakfast? Is there another eligible rate? Is the objection really price, or is the guest still uncertain about value?

In other words, the agent would diagnose before reacting.

That matters because keeping the booking direct should not mean throwing a cheaper rate at every guest who questions the price. Hotels have revenue strategies for a reason.

Sometimes the existing offer simply needs explanation. Sometimes another room-and-rate combination is more relevant. Sometimes an approved direct offer removes the objection. And sometimes the first rate remains the best option.

The commercial rules still belong to the operator.

The useful role of AI is to recognize what is happening in the conversation and apply the appropriate rule at the moment it matters.

A booking engine waits for a choice. A conversation reveals why the guest has not made one.

This is the important distinction.

A conventional booking engine presents inventory. Here are the rooms. Here are the rates. Here are the conditions. Choose one.

That works when the guest already knows what they want.

But a conversation contains another layer of information: intent.

“I need the cheapest option” means one thing.

“I need to be able to cancel” means something else.

“I saw it cheaper on Booking.com” is different again.

And “I like that room, but €190 feels too high” tells you more than a click on a rate card ever could.

Una reacts to that conversational intent.

The point is not simply that Una can display several offers instead of one. The point is that the conversation itself can determine whether another offer should be introduced at all.

If the guest has already made a clear choice, there is no reason to complicate the journey with more options. If the guest is still comparing, then presenting the relevant alternatives may help move the decision forward.

That is fundamentally different from a static booking engine showing every available rate and leaving the guest to figure it out.

More offers are not the goal

Imagine the guest has heard about two relevant options during the conversation.

One is flexible but more expensive. Another is cheaper but comes with different conditions. Perhaps there is another eligible room-and-rate combination that better matches what the guest has just said.

Without a connected workflow, the conversation can quickly become a memory test.

Which one was refundable? Which room was €170? Was breakfast included in the first one? Which link opens which offer?

The obvious response is to start comparing again somewhere else.

And somewhere else is often an OTA.

With Una’s Apaleo offer-comparison workflow, the alternatives discussed in the conversation can be carried into one direct path. If the guest has chosen, Una can take them toward that specific offer. If they are still deciding, the relevant options can be presented together so they can compare and continue from there.

The feature is multiple offers in one link.

The benefit is a clearer way for the guest to compare room types, rates and conditions.

But neither of those is the real reason an operator should care.

The business outcome is this:

a guest who hesitates over price gets another path to conversion before leaving the hotel’s direct channel.

The danger is not that the guest compares. It is where they compare.

Guests are going to compare prices. No booking technology will eliminate that behavior.

The more useful question is whether the hotel can keep the comparison inside its own sales journey long enough to remain part of the decision.

Once the guest leaves to reopen Booking.com, Google Hotels or another property, the hotel has introduced more competitors, more offers and more opportunities for the booking to disappear.

That is what makes the last part of the direct funnel so commercially sensitive.

The hotel has already done the expensive work of attracting the guest. Losing them here does not simply mean missing a reservation. In some cases, it means watching the same demand return through an intermediary — with commission attached.

The objective is therefore not to prevent comparison.

It is to avoid unnecessarily sending the guest elsewhere to complete it.

This is where “AI that acts” becomes commercially relevant

There is a large difference between answering a reservation question and continuing a reservation workflow.

A chatbot can answer:

“What rooms are available?”

It can answer:

“How much is the flexible rate?”

It may even correctly answer:

“What is your cancellation policy?”

But if the guest then says, “That is too expensive,” the next move is no longer an FAQ response. The system has to understand what changed and decide which permitted action should follow.

That is where Una’s Digital Employee model matters.

The AI is not there simply to produce a more natural sentence. It can use the conversation to understand intent, work with the available booking options and continue the workflow within the rules defined by the hotel.

The same logic can operate in supported text conversations, website voice and phone calls where the Apaleo booking flow and required link delivery are enabled.

For the guest, the channel should be irrelevant. They do not care whether the hotel internally calls something a widget interaction, a phone call or a booking-engine session.

They care that the hotel understood what they just said.

Humans already know how to rescue these bookings

Experienced reservation agents have always worked this way.

They listen for uncertainty. They clarify what matters. They know when another option may help and when reducing the price would be unnecessary. They can tell the difference between a guest looking for a bargain and a guest who simply needs more confidence in the offer.

The challenge is delivering that behavior consistently.

At 2 PM, an experienced reservations employee may save the booking in thirty seconds. At 2 AM, the same call may reach a team member dealing with arrivals, an outsourced line, or no one at all.

Across a multi-property portfolio, those moments multiply quickly.

That is why the more interesting AI story is not staff replacement. It is operating capacity.

AI handles the repeatable, time-sensitive parts of the workflow: recognizing the situation, applying approved rules, retrieving relevant options and keeping the booking moving. Humans remain focused on exceptions, sensitive negotiations and cases where judgment genuinely adds value.

More conversations can be handled without requiring workload to increase linearly with every additional property or guest interaction.

A better way to test reservation AI

If you want to know whether a reservation AI can contribute to direct revenue, do not test it only with the easy questions.

“Do you have a room next Friday?” proves that it can retrieve availability.

Try the sentence that puts the sale at risk:

“I can see it cheaper on Booking.com.”

Then watch what happens.

Does the system recognize that the guest’s intent has changed? Can it understand whether the comparison is actually equivalent? Can it apply the hotel’s approved rates and conditions? Can it offer another relevant path without automatically discounting? Can it keep the guest moving toward a direct reservation?

Or does it simply repeat the answer it gave thirty seconds earlier?

That is a much more useful benchmark for hospitality AI.

Because the value does not appear when the guest obediently follows the happy path. It appears when they hesitate just before conversion.

Don’t pay to reacquire demand you already had

OTAs are exceptionally good at converting demand. That is one reason hotels use them.

But if a guest is already inside the hotel’s direct funnel, the commercial priority changes.

The hotel has already acquired the attention. It has already started the conversation. It has already created a credible path to booking.

At that point, every unnecessary exit from the direct journey creates a new chance for the guest to convert somewhere else — potentially sending the same reservation back to the hotel with an acquisition cost attached.

So the objective is not to show more rates.

It is not to make the comparison screen prettier.

And it is not to give every price-sensitive guest a discount.

The objective is to keep more high-intent demand in the hotel’s direct channel long enough to convert it.

That is the business story behind offer comparison — and the standard against which the feature should be judged.