Are you also afraid of overselling? You’re not alone
By Ira Vouk, Co-founder of iRates
As we know, overselling (or overbooking) is a technique used in Revenue Management to offset anticipated cancellations and no-shows (wiki has an article on this subject). In other words, if you expect 2 cancellations and 1 no-show - you oversell by 3. That's the optimal behavior that maximizes revenue. Pretty simple, right?
But nonetheless, not very many hoteliers embrace this practice. It's still very common for most managers (especially in the middle tier segment) to close out availability on all channels before they reach 100% occupancy mark for a certain day. In most cases, this decision is dictated by the fear of dealing with this:
However, correctly implemented overbooking practices will minimize the chance of a walk while leading to a noticeable increase in Revenues (as well as profits). One doesn't need to have a lot of Revenue Management experience or knowledge to be able to achieve this goal.
In this article, I will try to describe some overbooking techniques and best practices. My goal is to show that overbooking is definitely underestimated. It is also not as scary as it may seem and leads to great revenue increases if done right.
Two types of overbooking techniques
As mentioned above, the scary image of a walked guest is generally occupying the minds of those who strongly disagree with the overbooking policy. Most popular mistake that leads to walks is picking the wrong time to overbook. As mentioned above, overbooking is designed to offset cancellations and no-shows. Let's ask ourselves: what is the main difference between those two occurrences? It is timing.
- cancellations can happen at any point in time, starting from 56 weeks before arrival (standard allowed lead time for transient bookings) until the end of the cancelation deadline
- no-shows always happen on the last day
Thus, we need to separate 2 different overbooking techniques: those that address cancelations and those that address no-shows. The latter is easy: calculate the anticipated number of no-shows and overbook on the last day by that amount of rooms. There are a number of good articles written on this subject (for example, "Overbooking ratio step-by-step" by eCornell).
But! What do we do with cancellations? There are articles explaining how to calculate average anticipated number of cancelations but no one tells you WHEN to do this, at what point in time (a little but a very important detail).
Timing is the key
Let's say you're planning your overbooking strategy for the 4th of July this year. You looked at your last year's performance report for the same day and discovered that you had total of 15 cancelations. That's the final number but it doesn't describe at what point in time those cancellations happened. If you're doing this exercise 3 days before arrival and you've just allowed the hotel to sell 15 rooms over capacity, chances are, you will be in this situation:
…simply because you didn't take into account the timing of your cancelations.
So here's the key: it is not enough to just calculate the total number of rooms to oversell. You need to build a curve describing the forecasted number of potential cancelations at any point in time.
Or, if you don't have any automated Revenue Management tool that would help you with this, you can use a simple excel spreadsheet. Something like this:
Note: if you don't have an RMS at your hotel, the forecasted number of potential cancelations can be manually calculated as an average of actual cancelations from similar days in the past.
Thus, in this example, you shouldn't worry if you find yourself overbooked by 5 rooms 10 days out. However, if you're looking at tomorrow's occupancy (1 day out), you should allow no more than 1 (unless you also anticipate no-shows, but that is a different exercise, very well described in the eCornell article referenced above).
Here's another trick: when anticipating cancelations, in order to maximize your revenue to its highest potential, overselling needs to happen at the peak of demand. This will ensure that:
- you sell those rooms at the highest possible price
- and (in some cases) you will still have plenty of time to wait for those precious cancellations
What do I mean by "peak of demand"? Guess what, demand is a curved line, it's not even for any day that you're selling in the future. Here's a good exercise: pick one day from last year (4th of July for example) and track the number of reservations booked per day, starting from 365 days before arrival.
You may get something like this:
Obviously, we're talking about high demand days only (when you expect to sell out) so there's only that many days that you would need to review for your property. You can use an excel spreadsheet to build this graph, if you don't have any automated Revenue Management tool that would help you do this.
The graph above is a courtesy of iRates LLC (www.i-rates.com) and describes the actual demand flow for the busiest convention in San Diego – Comic Con that brings about 130,000 visitors every year. The curve is similar for the majority of the city-wide conventions in San Diego.
You would notice 2 peaks: one about 140 days out, and the other one a few days before arrival.
It is also important to know that price elasticity (and thus, price expectation) is not constant either, it's also curved. For this event, the peak of the price expectation falls around the first peak of demand and then goes down:
So regardless of the fact that the second peak of demand is higher, the most profitable bookings for this event can be captured 3-4 months out. This is when the majority of the reservations need to happen in order to maximize profits.
I need to note here that if you've planned your pricing strategy correctly, you won't often find yourself in a situation when you're close to 100% occupancy that many days out. However, these situations do happen when demand greatly exceeds supply and you have reached the ceiling of Rack rate allowed at your hotel, so you can't increase your rates any further (for various reasons, one of them being price gauging). In this case, you implement other non-pricing Revenue Management techniques (which is a whole separate discussion) but even then you may end up selling out in advance. The situation shown in the graph above is not very common and is just being used as an example. In your case, the peak of demand may fall on 7 or 5 days out for the majority of high demand events, or maybe always on the last day. Each hotel has its own unique demand structure and you would need to analyze it.
In any case, the general rule is to oversell at the peak of demand in order to maximize your profit. Do this and enjoy looking at those opaque 40% off bookings (those that you never meant to allow but somehow they always slip through) turn into high-profit Rack rate reservations from direct customers.
Here's another example for better understanding.
Below is an excerpt from an excel spreadsheet that describes the booking pace for an event.
- Manager 1 doesn't like to oversell and employs Scenario 1
- Manager 2 understands overselling techniques and uses Scenario 2
Let's review the numbers.
As you see from the spreadsheet, the peak of the demand and price expectation falls on the 11-9 days before arrival. Under scenario 1, the manager closes availability on all channels during the 3 days, unwilling to oversell. This leads to total of 10 denials that could have been booked at the premium rate. After that, when demand decreases (this may happen due to various reasons, i.e. convention hotels releasing their group blocks thus increasing the supply in the market, or competitive hotels lower their rates pursuing full sell out), booking pace slows down and cancelations start coming in. Thus the manager is forced to respond to the slowing demand conditions with the lower rate, in order to not end up with too many empty rooms. Manager aims towards full occupancy and is able to achieve it at the end, if not for the 3 no-shows. As a result, 3 rooms remain unoccupied. One doesn't need to have a pretty Revenue Management certificate to understand that this strategy is far from optimal and leads to significant revenue losses.
Let's review scenario 2. In this case, the hotel is not closed out at the peak of demand and total of 10 rooms are booked over capacity at the premium $239-$259 rate (i.e. denials turn into actual reservations), anticipating future cancelations and no-shows. The strategy allows achieving 100% occupancy as a result. What we also see is that under this scenario, the manager is not forced to lower the rate to reach full sell out. Moreover, he even increases it further as the demand stays strong, and accepts the denials from the first scenario. The rate remains at $259 until the day is closed.
Estimated amount of lost revenue under Scenario 1 vs Scenario 2 is about $1750 (considering cost of empty rooms and lower ADR of the latest reservations). And this is just for 1 day. For a 4-day event, multiply this number by 4 – and you get $7000 of pure loss (that goes straight to the bottom line) from incorrectly managing your overbookings! Multiply this by the number of events during the year. This could add up to a 6-figure number. Another note: example above is taken from a real 100-room middle-tier property. For a larger hotel, the gains/losses would be proportionally greater.
To summarize, I would like to emphasize one more time that the effect of incorrect overbooking policies (or the lack of thereof) is being widely underestimated in the industry, especially in the limited service/middle tier segment. Proper overselling at the peak of demand helps hoteliers sell their rooms at the premium rate and not leave money on the table from empty rooms due to anticipated cancelations and no-shows. This is optimal strategy that leads to revenue (and profit) maximization. The additional revenues gained from the overselling techniques go straight to the bottom line.
Ira Vouk, CRME
Co-founder, iRates LLC.
iRates is a new-generation software-as-a-solution Revenue Management system that was developed specifically for limited service and middle-tier hotel properties. The tool was designed with these types of properties in mind. It addresses unique characteristics and requirements of this segment, based on years of experience in this segment.
The extremely user friendly interface along with complete automation of the most important revenue management decisions allow even a junior level employee with no experience to drastically increase production and improve RevPAR. iRates employs the newest adaptive algorithm based on the theory of reinforcement learning that automatically adapts to each hotel's local market conditions and re-adjusts the strategy according to the demand fluctuations, with the goal to maximize the bottom line profits.The algorithm has been successfully applied to other industries but is unprecedented in hospitality.