Revenue Management Is About Probability

The short answer

There is no single correct nightly rate, only a price and a probability of booking before the dates expire. Good revenue management weighs what a rate would earn against how likely it is to sell, then updates the decision as booking pace, visibility and the calendar change. A disciplined process makes each choice more informed; it can’t promise that a particular night will sell.

There is no correct nightly rate. There is only a price and a probability of booking.

Will a particular stay sell at $450 per night? Possibly. Would it book faster at $400? Perhaps. Would $350 attract a reservation that otherwise never arrives?

You cannot know those outcomes in advance. You can make an informed judgment, observe what happens, and update your decision as the evidence changes. That is the work of revenue management.

At Host Tahoe, I view pricing as a decision made under uncertainty. The objective is to make the strongest decision the available information supports, then keep learning while there is still time to act.

A rate recommendation is a forecast in disguise

When someone recommends $450, they are implicitly making assumptions about demand, competition, the home’s appeal, and the time available to find a buyer. The number can look precise even when those assumptions are uncertain.

A holiday, an attractive view, or a strong review history may support a higher price. None guarantees that a suitable guest will arrive, find the listing, and choose those dates at that total.

The opening statement doesn’t mean every rate is equally defensible. Some prices are clearly better supported than others. It means there is no single, permanent answer waiting to be discovered independently of the stay, the market, and the moment you make the decision.

The useful question is:

At this price, how likely is a suitable booking before these dates expire, and what are the consequences if we wait?

Price and booking probability have to be considered together

Consider a hypothetical two-night gap approaching arrival. Suppose we could estimate the probability of selling both nights before they expire at three different rates:

Nightly rate	Two-night lodging revenue	Assumed probability of booking	Expected lodging revenue
$450	$900	30%	$270
$400	$800	50%	$400
$350	$700	70%	$490

These probabilities are invented solely to illustrate the decision. They are not Host Tahoe forecasts, market benchmarks, or measured results. The example assumes the rate stays unchanged until the dates expire and that the gap either sells in full or earns nothing.

Expected revenue is the booking revenue multiplied by its probability. Under these particular assumptions, $350 produces the highest expected lodging revenue even though it has the lowest nightly rate.

The property will not actually collect $490. It will collect $700 if the gap books, or $0 if it remains empty. Expected value is a way to compare uncertain options, not a promised payout.

Different probabilities could favor $400 or $450. Booking costs can change the comparison too. For an owner’s return, the more useful calculation considers what remains after the additional costs of hosting the stay.

The point is to evaluate the likelihood of collecting a rate alongside the amount you would collect.

Most of the time, the probabilities are uncertain too

An operator rarely has enough evidence to say a particular stay has exactly a 70% chance of booking. A small portfolio, shifting demand, and differences between homes make precise estimates difficult.

That doesn’t make probability-based thinking useless. It makes honest judgment essential.

You can ask whether the evidence supports a relatively strong or weak chance of selling, whether that chance appears to be improving, and how confident you are in the assessment. You can also ask whether the decision would change if your assumptions were a little too optimistic.

If holding a premium rate only makes sense under an unusually optimistic forecast, the decision is fragile. If a modest adjustment looks reasonable across several plausible outcomes, there may be a stronger case for acting.

Confidence should reflect the quality of the evidence. A precise-looking number cannot substitute for that evidence.

New information changes the decision

Imagine an open weekend six weeks away. The property has historically attracted guests closer to arrival, relevant searches show useful visibility, and comparable homes remain widely available. Holding the rate may be reasonable.

Two weeks later, the dates are still open. That fact alone doesn’t settle the question. Is the property behind its usual booking pace? Are guests reaching the listing? Has the competitive offer changed? How much of the likely booking window remains?

Then a neighboring reservation creates an isolated two-night gap. The opportunity changes again: those dates can no longer support a longer stay.

Each observation changes the information available for the next decision.

I consider booking pace, guest-facing totals, observed search visibility, inquiries and conversion where available, local demand, competing offers, and the shape of the remaining calendar. I also check the reliability of the evidence. A competitor’s blocked calendar does not, by itself, prove a paid booking or reveal the rate collected.

Updating a recommendation when the facts change is part of disciplined management.

Waiting preserves an opportunity, and consumes time

The simple probability example holds each rate constant. Real calendars allow a more complicated choice: hold $450 today and reassess later.

That flexibility has value. Accepting a lower-priced booking now closes the possibility of selling those dates to a higher-paying guest later.

Waiting also has a cost. Guests continue making decisions, and the inventory moves closer to expiration. A lower rate offered later may no longer reach the same pool of potential buyers.

This is the central tradeoff: the possible benefit of a better future reservation versus the risk of losing a worthwhile opportunity now.

It deserves different judgments in different circumstances. An intact holiday week months ahead may warrant patience. A weak two-night gap approaching arrival may warrant action sooner.

Reducing a rate in stages is therefore a strategy to evaluate, not a guarantee that you can always secure a booking at the last minute.

The calendar makes the probabilities interdependent

A guest usually buys a sequence of nights. Accepting that reservation changes which future reservations remain possible.

A short booking in the middle of a peak week may prevent a longer stay. A weeklong reservation may fill weekdays that have little chance of selling separately. A two-night gap between confirmed stays has a much narrower set of options.

This means the best decision cannot always be found by estimating each night in isolation. You also need to consider the likely booking patterns and what accepting one stay does to the value of the remaining inventory.

The question becomes:

Which decision gives this calendar the strongest expected return, given what we know now?

A good decision can still have a disappointing outcome

Suppose holding a higher rate was well supported by demand and booking history, but the dates remained empty. That result deserves review. It doesn’t automatically prove the decision was irrational.

The opposite is true too. A high-priced booking arriving at the last minute doesn’t establish that waiting was the best repeatable strategy.

There is also a counterfactual you cannot directly observe: what would have happened at the other price. Once a stay books at $350, you cannot know with certainty whether it would have sold at $400.

This is why I record the evidence, the action, and the reasoning before the outcome arrives. Over time, we can examine whether similar decisions produced the results we expected, while accounting for differences in season, lead time, and available inventory.

One successful booking is useful information. A consistent pattern across comparable decisions is stronger evidence.

Hospitality Intelligence makes uncertainty easier to manage

Hospitality Intelligence is being developed around this approach: connect the relevant evidence, make assumptions visible, identify changes that deserve attention, and help the operator decide whether to hold, watch, or act.

The goal is better judgment supported by an ongoing record of decisions and outcomes. Technology can organize more information; the recommendations still need to account for the property, the guest, the calendar, and the owner’s objectives.

No system can promise that a particular night will sell at $450, $400, or $350. What a disciplined process can do is make each choice more informed, and recognize when yesterday’s reasoning no longer fits today’s opportunity.

Frequently asked questions

What is expected revenue?

The booking revenue multiplied by its probability. In the invented example in this article, $350 a night for a two-night gap at a 70% chance gives $490 of expected revenue, more than $450 at 30% ($270). The property never actually collects $490; it collects $700 or $0.

If a lower rate books, does that mean the higher rate wouldn't have?

Not necessarily. You can't observe what would have happened at the other price. That's why it helps to record the evidence, action and reasoning before the outcome arrives, so similar decisions can be compared over time.

Should I lower rates in stages as arrival approaches?

It's a strategy to evaluate, not a guarantee. Waiting preserves the chance of a better booking but consumes time, and a lower rate offered later may no longer reach the same pool of guests.

Written by Eric, founder of Host Tahoe in North Lake Tahoe: more than 25 years in hospitality and nearly a decade operating vacation homes.

Revenue Management Is About Probability, Not Certainty