November 1998 – The marketplace is becoming more complex and competitive than ever, and everyone’s looking for ways to increase revenue and profitability. Given that revenue management systems are routinely claimed to return 3-6% revenue increase for their users, why isn’t everyone flocking to buy them?
Even in good times, the best properties look for the highest return on their investment in rooms. When things get tighter – as they always do – it’s far more financially productive to stay focused on maximizing revenue than to try to survive solely by cutting costs to the bone. The best rooms revenue results come from correctly forecasting exactly how many of each room type on each night can be sold to the highest-paying customers. In today’s increasingly fragmented market, though, doing this manually can become impossibly complex.
The salvation you need may just be a property-based revenue management system (sometimes called a yield management system), that can calculate for you. Several systems are available specifically designed for the hospitality market; each vendor claims that its system will produce a 3-6% increase in revenue, and achieve a consequently fast payback. Nevertheless, many properties are reluctant to take the plunge and use one. Why?
What’s RM All About Anyway?
This isn’t going to develop into a detailed treatise on the theory and practice of RM. There are plenty of better sources for that, starting with Bob Cross’ excellent book, “Revenue Management,” which grew from his experiences at Delta and his founding of Aeronomics (now Talus). But it may help to understand the real value of the systems if the basics are covered first.
In a nutshell, revenue management is the science of using past history and current levels of booking activity to forecast demand as accurately as possible to maximize revenue. Everyone attempts to even out the peaks and valleys of occupancy by increasing rates during times of high demand and discounting them when demand is low. Most properties break this approach into manageable chunks by applying different rates and booking constraints to different market segments, and make changes as necessary as bookings come in. Too often, however, the trigger for action, especially discount action, is an emotional reaction to low levels of bookings for short-term arrivals, and not a reasoned decision based actual – but hard to see – trends.
However, if a property has good historical data on its major market segments and their typical booking patterns, it might be less tempted to offer a discounted rate if it could reasonably expect that full-rate guests would, as they have done before, book only in the last thirty days before arrival. Equally, if it knew it had a pattern of medium-stay guests (say, four to five nights) at a particular point, it might be more confident about turning down a full-rate single-night stay in favor of the greater overall revenue available from the longer stay, even at a slightly lower rate. And if it knew it had a small but reliable group of guests who regularly called for a room right at the last minute and were willing to pay a premium for that flexibility, perhaps it could increase its rate for those guests without driving them away.
This approach to managing rates and length-of-stay restrictions to get the most income over any particular period isn’t rocket science. Given appropriately-detailed records (enough market segment divisions to be useful, not too many to be monitored effectively) this level of Revenue Management is easily done in a manual environment, and falls under the general – but crucial –heading of knowledge-based decision-making.
Every room in a property on any night has a wide range of potential customers. For maximum profitability it’s vital to know how much someone who really values that particular room on this date with this amount of lead time is willing to pay for it, and then to have the courage and confidence to hold out for that rate from that customer.
Where do you get that confidence? From the knowledge that you have a firm bedrock of property history, coupled with your own assessment of how similar current circumstances are to those of the historical data, and a clear understanding of the purchasing patterns of your particular customer mix. In other words, your Revenue Management allowes you to make informed decisions, not just relying on gut feelings.
How RM systems help
The above RM practices have become essential to effective, focused marketing. But not only has the marketplace become increasingly fragmented and fast-paced, with each subdivision and micro-segment having its own judgements on the relative value of a room, outside influences change rapidly too. The constant re-calculation of the value of taking a particular booking, compared to all the other possible bookings you might expect to be requested for that room later, becomes an impossible mental task.
But this kind of high-speed evaluation is where computer systems, RM systems in particular, excel. Far faster than even the most experienced director of sales, an RM system can take historical data, current booking levels and the forecasts of which market segments are expected to be looking for these rooms in the future, balance the various rate tolerances of these segments, and set minimum rate hurdles and length-of-stay constraints below which you should not accept a current booking request. These forecasts are updated at least every night – some systems do it every hour – to make sure that there’s no delay in deciding whether to take a booking; the criteria are always right there on the users’ screens.
They also refine their calculation formulae if the original assumptions don’t seem to be working. To take an example from Opus2’s Eric Orkin, you normally assume that a tossed coin has a 50:50 probability of coming down heads or tails. If it comes down heads seven times in a row, you’d expect the eighth toss to have a high probability of coming down tails, and could make predictions based on that. But if it comes down heads three more times before showing tails, perhaps you’re not dealing with a symmetrical coin; you might want to skew your formulae a little towards the actual characteristics you’re recording, say to 60/40, and then continue to watch developments.
Rather than be totally occupied with these kinds of calculations, management can leave them to the system and stand back, monitoring the activity and checking that the overall marketing direction is still appropriate. They can ensure that constraints and special-event knowledge that the system wouldn’t otherwise be aware of (an upcoming city-wide conference, a new sports franchise’s schedule, a new hotel opening nearby) are entered appropriately, and that system parameters are updated to reflect what’s happening in the real world as bookings come in. The RM system can highlight potential problem areas that fall outside pre-set limits, allowing the managers to focus on these exceptions.
Note the critical point here; the systems don’t make your marketing decisions for you, they just try to achieve the goals you’ve set in the context you’ve defined through experience. Humans can’t calculate as well as computers; computers can’t make expert evaluations of the effect of changes in their operating environment. Setting your marketing goals and strategy, and changing them if you’re not reaching the right potential guests, is still something you have to do yourself.
Setting these baselines is another area where automation can help out, of course. Guest history and database analysis tools allow a far more detailed look at historic and potential customer base, and automated marketing plan development and tracking tools (such as those available from Driving Revenue) can make a really significant difference to setting your focus and goals. Identifying what’s actually happening from your RM results can also provide valuable feedback. The most productive approach is to let humans and computers each do the work they do best.
Who can use these systems?
Pretty much any property can benefit from a RM system. Size isn’t so much a factor – successful implementations exist in properties from 40 rooms to 1,200 – but typically the more complex the environment, the more varied the mix of market segments, and the greater the peaks and valleys in current occupancy, the greater the benefit. A high percentage of transient guests also helps, since applying RM principles to group bookings is still a relatively young art.
It’s easier to say who wouldn’t get much benefit. It’s a short list:
1- hotels with a very high percentage of group business at pre-negotiated rates with guaranteed last-room availability, i.e. if a single room is available at rack rate, a group member can take it at the discount. Even then, RM can help by shifting stay patterns, e.g. locking out that available room-night by imposing a minimum-length-of-stay at that point.
2- a hotel which sells all its rooms, all the time, at rack rate – and they need a consultant to advise them on re-positioning their marketing strategy!
So why aren’t more hotels buying? We can’t afford it
The price issue has a couple of aspects, one illusionary and one more real. The illusion is the belief, coming from most people’s experience with RM coming from dealings with the airlines, that these systems only apply to hugely complex operations and must therefore be hugely expensive. Maybe that was true once, but now the programming expertise and the power of those mainframe systems has migrated down to PCs. Hospitality-specific systems are available from XXX and YYY per room, or about $25,000 for a 200-room property.
At least one vendor – IDeaS – offers a revenue-sharing approach: no revenue gain, no payment. While determining how much of that gain is directly attributable to its system pretty much has to be done with their own measurement tool – unless you’re in an unchanging market and have kept your rates constant before and after RM installation – those hotels that have gone this route do seem to validate the systems value. If a decision to purchase one outright is made, even a 3% revenue increase should see a ROI pretty quickly – but you’ll need to buy it for the right reasons and use it in the right way, of course.
The pricing issue that’s real is the one of every property having multiple projects competing for the same capital dollars, and, as for all computer-based systems, it can be hard to see any immediate tangible benefit from investment in an RM system. You can’t feel higher revenue like you can feel new guestroom furnishings or a new lobby carpet – although wouldn’t you like to invest in a system that will produce more income to spend on such items? This puts a premium on the realistic management of expectations, using relevant examples of similar properties’ favorable RM results, and on good, regular feedback to management of the actual impact of a system.
This feedback should include the regular reporting of revenue changes, with a realistic assessment of how much of each period’s improvement is due to market changes (general inflation) or hotel action (general rate increases) and how much to the system getting the maximum return from a constant rate structure. It can be generated either manually or via a tool such as IDeaS’ before-and-after comparison program. But there have been enough successful implementations of RM systems to leave no doubt that they have a real and positive effect. As always, good communications is the key to getting the benefits appreciated and more widely known.
I don’t understand it
As for being too complex to understand, the airline origins of revenue management do seem to have given it a mythic image, of being beyond the ability of mere mortals to understand or to manage. Surely it’s too much for an individual property? Not at all; it’s the individual property that’s best positioned to use these tools to leverage its own knowledge and experience, and fine-tune them to reflect local conditions. Further, a centralized hotel chain RM operation only really works when the CRS sees all the rooms available at all the properties, since local input on local context is vital.
Certainly the statistical math behind computer-based RM systems is complex, requiring rocket-scientist levels of IQ to develop and understand. But the calculations required to build a car are also complex, and we don’t expect drivers to be as technically-gifted as their vehicles’ designers. You can run a sales operation manually with good and appropriate goals in view, just as you can walk to any destination you choose. But you can also learn basic RM system operations, and learn to trust that the machine will do what you tell it, just as you learned to drive. You’ll reach your goals faster and more effectively – and avoid the risk of getting run over by the traffic that’s already on the road.
Who’s in control here?
Finally, RM systems don’t control you; you control them. They don’t make your marketing decisions for you; you still have to tell them where to take you, and what special factors to allow for.
Some managers are more comfortable just working with the rate management functions built into many modern property management systems; sell only five of this package, cut off these rates if occupancy goes above 75% on these days, and so on. They understand exactly what the system’s doing because they know what they told it and can see all the steps involved. With a full RM system, they have to trust that the “black-box” nature of its calculation engine is doing the right thing, a premise that can only be proven by checking the results. This trust and confidence will only come in time, but it will come, as the accuracy of the forecasts is borne out. If you don’t like the results, resist the temptation to blame the system; it’s only doing what you told it. Re-examine your strategy, and see if it’s still appropriate for what’s actually happening in the real world.
Criteria for success
So what does it take to make a successful RM implementation? While all the vendors agree that hotels of all sizes and types can benefit from their products (no surprise there), they also all agree that the criteria must be met for the systems to work effectively. These are:
A- The right culture
B – Management commitment
C – PMS interface
D – Good, plentiful historical data, and
E – Constant attention to relevancy and context.
The right culture
Like any other computer-based system, RM systems are tools, not magic bullets. Dropping a PMS into a disorganized rooms operation just gets you more disorganized faster. Dropping a RM system into a sales operation running on instinct and guesses will give you misleading results you’d rightfully have little confidence in. The most important requirements for a successful RM implementation are that the property must already believe in the operational importance of knowledge-based decision-making (or be willing and ready to embrace it), and must have a focus on maximizing revenue rather than minimizing cost.
One reflection of the culture would be the nature of any incentive plan for your reservations agents. If bonuses are based on the number of room-nights they sell, you may increase occupancy but at what rate? The temptation to offer discounted rates during peak demand periods could be too high. A revenue-focused organization would lean more toward rewarding agents for closing increased RevPAR sales in high, medium and low demand periods.
As a further benefit, encouraging awareness of the property’s unique value and revenue potential rather than constantly reminding everyone how tightly costs must be controlled has an enormously energizing effect on morale and productivity, even before you leverage that energy with a RM system.
The enthusiasm and direction must come from the top, and exist throughout the operation, to get the most benefit from the system. Everyone needs to accept that it’s a tool to be used, kept sharp and its strengths and weaknesses understood and worked with. Ideally each property should have a RM champion, someone who focuses on keeping the systems parameters current and revenue flow maximized across the property. A lot of time and resources will be invested to set it up. Stay committed to making it work and understanding what it tells you, and you’ll see results.
This sounds as if it should come first; after all, the very basis of an effective system is the wealth of detailed reservations data captured by you’re the PMS, and without an automated interface to transfer it to the RM system you’ll never keep up. However, it’s still essential that the property’s culture accept the need for the concepts first, otherwise no RM system will have a chance of being used properly. The need for a PMS interface may place limitations on the options somewhat, since interfaces are produced according to customer demand. Most RM systems have interfaces with varying degrees of complexity to at least one of the major PMSs (Fidelio, Geac, LANmark, etc.). (Check with your PMS vendor to see not only if it has an interface to transfer data to an RM system, but that it can also make maximum benefit of the rate hurdle and length-of-stay information coming back.)
Good, plentiful historical data
The accuracy of any RM systems forecast is directly related to the size and detail of the database it is working from, regardless of the specific mathematical approach it takes. Ideally, most vendors like to have at least a year’s worth of complete, detailed property data to use as the forecasting basis, usually from the PMS’ archived end-of-day back-ups and reports. Frequently, however, this amount of data isn’t available, or isn’t as complete as it could be. For example, some PMSs don’t track all changes to a group block along with the date and time they occurred, and it really helps to know what booking constraints were in effect when a reservation was taken. In such cases – and especially with newer properties – the available data has to be supplemented with general information from management experience, and with additional system detail captured on-site. Most vendors will record live data for three months before cutover to give a more accurate starting point.
While systems’ accuracy improves from day one as real-world data can be compared to predictions, it’s important to have the baseline be as solid and complete as possible.
Constant attention to relevancy and context
Always keep in mind the basis of the RM forecasts when considering the results. Soon after cutover, forecasts will still be based on some unvalidated assumptions and incomplete data, and maximum forecasting efficiency usually comes after a full year of operations, when all real-world seasonal variations have been included.
Also remember that RM predictions are based upon the parameters plugged in. Despite the system’s attempts to compensate for unusual levels of activity – high or low – there is no way it can detect changes in context, such as a competing property opening across the street or announcing a massive discount incentive program. As with a manual marketing strategy, always keep an eye on its current relevance; the playing field may have changed.
But not too much attention…
On the other hand, micro-managing the system and second-guessing it can be deadly to results. You’re working in a complex, fast-paced environment where situations and influences change by the minute. You’ve bought a computer system to handle the monumental task of recording and evaluating the millions of pieces of data that affect this. Trust the answers it gives you. You don’t have time to check its arithmetic, and it probably knows rather more about statistical probabilities than you do. Work with it, not against it.
Humans don’t seem to be genetically programmed to understand statistics. Scientific research has some interesting things to say about the success rate of using your gut reaction and instincts to override computer-based decisions (see sidebar). Certainly you have to keep an eye on the big picture, but you gotta have faith, too.
A Word on Ethics
Sometimes you hear comments that the very nature of RM is somehow unethical, that properties have no right to take such advantage of people’s last-minute needs by charging them extra, or that the very pursuit of every possible dollar is somehow immoral.
But most of us have trouble fully understanding the value that someone else can put on our product under circumstances that don’t apply to us. That doesn’t mean that it’s not a fair price; it becomes a fair price to that person at that time in those circumstances. Understanding the effects of these value judgements is what RM systems do well, by impartially evaluating all the circumstances under which such guests have paid high rates and predicting when they’re likely to do so again. If it sells, it was the right price for someone. You don’t have to understand it, just believe it. As Bob Cross says, when a last-minute guest arrives to take your last room and complains that the price is pretty steep, your staff can say, with perfect confidence and a smile, “Sir, we’ve been expecting you, and we kept a room for you. If it was priced any lower, it wouldn’t be here.”
Nevertheless, an obsessive focus on making every last dime can be a pretty unhealthy approach to life. Yes, you need to get the best return you can for your operation, but sometimes that return doesn’t come in monetary form. Airlines discount last-minute fares for bereaved passengers; hotels often offer cheaper rates to the families of critically-ill children. An RM system doesn’t force you to be Scrooge; you control its guidelines, and you can make exceptions. Combining good business judgement with compassion will give you a better overall reward – and can also pay off later in terms of increased business from an enhanced reputation.
In good times and not so good, any hotel focusing on increasing its revenue by every means possible will do better than one solely concerned with containing costs. If you’re not already following revenue management principles, think hard about how long you can stay competitive without them. And if you are, a RM system will take away the sometimes overwhelming burden of keeping up with all the essential forecasting details, and give you the time and confidence to apply your experience where it’s most effective – in the big picture. Go for it.