R.I.P. R.P.I.

All opinions are those of the author

The Achilles Heel of Hospitality

Nearly every industry has developed a set of standards for its investors to evaluate the performance of companies against industry averages. The industry often called “mankind’s second oldest profession” may also have the most antiquated measurement processes. The competitive performance figures used in hospitality today were designed before the terms “big data” or “analytics” existed. The study below will explore the background, the mathematical errors, the misinterpretations, and the opportunity cost from an obsession with a faulty metric. May RevPAR Index rest in peace.

Fax It to Me

In the 1980’s and 90’s, hotels in the U.S. would do “daily shop calls” where neighboring hotels would call each other to share occupancy figures from the night before. This rudimentary process was a tiny footnote in the whole of a performance management process. In most cases, it was the graveyard shift at the front desk tasked with making the calls. Over the years, hotels exchanged more information — the prior night’s occupancy and a forecast for the next few days. Hotel operators mostly used the insights to know where to relocate guests when oversold.

Pricing and the Average Daily Rate (ADR) figures were rarely exchanged, but even while participants kept on the correct side of antitrust concerns, the process was highly manual and raised eyebrows in an era of rapid technological innovation. Smith Travel Research was providing monthly aggregate summaries, but those reports (like today) were outdated by the time they were released three weeks after the final day of the month. The need for faster data to enable faster decision making opened the door for the daily reports, called the DaySTR, where major hotel company headquarters began sending large data files with occupancy, ADR and RevPAR for all their hotels to a central repository in Tennessee for the data to be aggregated and pumped back out to hotel managers through email — a cool new messaging system.

At a time when people happily waited three minutes for a song to download through dial-up internet, there was nothing alarming about this part of the hospitality business. Since the dawn of the internet though, virtually nothing has changed in hotel competitive reporting.

The Math Does Not Work

Being imperfect may not be reason to kill such an entrenched metric. RevPAR Index however is not simply flawed, too often it is not even directionally accurate.*

1. Hotel size distorts performance realities. As analysis by All In Ltd. shows, equal performance is not fair share. Revenue divided by capacity is flawed, so when two hotels equally capture 50 of the 100 customers in a market, the “market share” would be the same, however when divided by capacity the larger hotel appears to have lost.
2. An aggregate index from indexes is flawed.
3. Hotels that appear in the comparable set of multiple hotels from the same brand are over-counted in brand aggregates. In a market where five Hilton hotels each have the same Hyatt in the comparable set, that one Hyatt will count five times in Hilton’s aggregate performance.
4. Aggregate brand level results always include some of their own brands within each other’s comparable sets. While the Holiday Inn brand does not compete with itself, a Holiday Inn legitimately competes at a local level with other Holiday Inns. As one grows its RevPar Index (RPI), it neutralizes any real gain or loss from the other when aggreged at a brand level.
5. Industry consolidation dilutes the relevance of RPI. As the biggest big brands grow their share of the total hotel supply, STR’s comparable set composition rules force the exclusion of true competitors reducing the relevance of RPI.
6. “Non-comp” is an exemption status afforded only to the internal brand aggregate measurement. This distorts an entire marketplace for all but the subject hotel. Any market with any hotel under renovation at any point in 2 years has inaccurate RPI changes (chart below).

The Interpretations of RPI are Inaccurate*

In addition to dubious math, the use of RPI in contracts between owners, management companies and lenders has created a set of entrenched misinterpretations and applications that the metric cannot support.*

1. RPI should not be confused with “market share”.
2. There is no rationale for expecting RPI growth annually or infinitely (particularly in a capacity constrained environment).
3. Competitive sets are too small to accurately represent dynamic marketplaces.
4. Competitive sets are arbitrary and often determined to produce a desired result.
5. Winning and losing has no meaning under Smith Travel Research’s year over year comparison mechanism. The same mathematic flaw illustrated with renovations applies to any non-repeating event. The “winner” of last year’s event is ensured to “lose” those dates the next year (chart below).
6. The impact of renovations and supply changes are poorly understood. The rules that protect individual hotel data ensure that there is insufficient detail for clarity leading to many misinterpretations.
7. The cost of customer acquisition is ignored. RPI encourages hotels to make poor profit decisions by “buying share” through high-cost intermediaries, kickbacks and expensive sales leads.
8. Ancillary revenue (restaurant, bar, spa, retail, etc.) is not taken into consideration.
9. RevPAR Index of 100 is an arbitrary and irrelevant mark; leaders and investors should not consider “100” as validation of anything.

* Assume capacity of 100 and ADR static all 3 years at all 4 hotels ** Assume Marriott internally removes its hotel as “non-comp”. *** Assume that displaced demand shifts from Marriott equally through the comp set

Focus is a Finite Resource*

The pain that COVID-19 has inflicted on the industry will exceed any of its disruptive benefits. The RPI metric is so ingrained in hospitality culture, compensation models and contracts that it goes unchallenged. However, the massive job losses will require management teams to become better focused and perhaps lose interest in a flawed metric.

Focus on RevPAR Index has distracted brand management companies to an astonishing level. Owners and investors are no richer for the millions of hours having been spent scouring STR reports. The opportunity cost alone is massive considering the time analysts place pouring over faulty figures and leaders waste explaining them.

If every hotel spends one hour a week analyzing, discussing and justifying DaySTR results, a company the size of IHG would dedicate more than 200,000 working-hours annually in the U.S. alone. One hour a week is a grossly conservative estimate considering the number of hotel-level staff involved in “market share” conversations. This estimate makes no consideration for the legions of highly-compensated regional and corporate resources also involved. Since the advent of the DaySTR report, over 20,000,000 hospitality-working hours have been dedicated to a faulty metric just in the U.S. by the big four brand management companies alone.

The industry can no longer afford to dedicate resources to anything but the most meaningful activities. RevPar Index obsession is an easy item to cut.

NUG is King**

Net Unit Growth (NUG) is the number of new hotels that a brand management company adds to its system. The big hospitality brands grow their companies not as much by growing RevPAR but by increasing their hotel counts. It is an industry standard for major brand management companies to take approximately 5% of every hotel’s revenue. Over a full economic cycle (2010–2020 for example), an average of 3% RevPAR growth is normal, so if a brand management company outperforms its peers by growing RevPAR by 5% instead of 3%, the company will improve its year over year revenue by $3,650 (orange). If instead, the brand management company adds another hotel that only performs to the market average, the company grows YOY by $182,500 (blue).

*All hotels assumed to be 100 rooms at 80% occupancy annually

Brand management companies that prioritize expanding unit counts will grow fastest. As the staffs of corporate headquarters are being gutted in the wake of COVID-19, it is reasonable to expect the portions of those companies dedicated to RevPAR growth to be reduced faster than the hotel development teams.

The company that grew from 1000 to 1100 hotels is making $2.7M more annually despite identical performance results. The RevPAR Index performance premiums between brands tend to move marginally, by tenths of a percent annually. Even if brand X outgrows brand Y in RevPAR by one full percentage point every year for an entire decade, brand Y will still be more profitable each year for having grown its unit count more aggressively. Net Unit Growth is everything.

Let the Games Begin

To grow NUG, brand management companies have development departments selling their brands to prospective owners. The pitch is that choosing their brand will yield the best ROI, but the measurement used to prove ROI is RPI. This is where the distractions begin. An owner once mused that development teams from all four major brand companies “proved” their case for delivering the highest ROI through demonstrations in which each showed superior “market share” results using RevPAR Index (RPI) figures. While this should be impossible, the math flaws outlined above allow each company to make this argument.

Once an owner selects a brand, the contract normally includes a clause that allows the owner to fire the management company if that asset falls below a certain RevPAR Index for consecutive years against a pre-defined competitive set of neighboring hotels. While hotels under major brand flags rarely fall beneath those pre-determined RPI marks, the real effect is that management companies allocate an irrational amount of attention on a silly metric; first to ensure that assets do not leave the system and second to so that the development department’s story for more NUG stays intact.

Everybody, Come See How Good I Look

It used to be that a hotel would renovate once each generation by closing its doors for a year, performing massive construction work and reopening to great fanfare. When wealthy individuals owned hotels, rather than REIT’s and investment groups, egos played prominently in decision making. This is still evident in Asia where the ultra-wealthy build the world’s most glamorous hotels as monuments to themselves. In the U.S. though, very few trophy hotels exist. The Grande Dames of yesterday are regularly gutted for any incremental dime.

The rise of institutional investment in commercial real estate and the evolution of asset management in hospitality shined a bright light on the financial losses from unrealized revenue that occurs by shuttering a hotel completely for renovation. It makes little sense to lose a full year’s worth of revenue by closing completely when the hotel could operate during construction at partial capacity by renovating several floors at a time based on projected gaps in occupancy. When a hotel is forecast to be 70% occupancy, construction teams work on the other 30%.

For owners, this drip-approach to investment is optimal, capturing the available revenue while methodically injecting smaller doses of capital. While financially smart, a downside is that hotels can become a patchwork of conflicting design concepts. Riding the elevator is akin to time travel where one floor speckles with its sleek and modern appeal, crisp lines and fake high-end furniture, while the next stop looks like an ‘80’s sitcom set.

RevPAR calculations rely on just three figures- revenue, rooms sold and capacity. When under renovation, the capacity figure is nearly always inaccurate. Inaccurately low capacity one year means inaccurately high capacity comparisons the next year. The constant renovation cycle means there is rarely a two-year period without a renovation among a group of competitors. Decades can pass without accurate comparisons in an entire market while hotel management teams are rewarded and reprimanded based on nothing more than the bad math that this cycle creates.

Management teams at hotels not under renovation rise and fall through no effort of their own while a hotel under renovation is likely labeled “non-comp” for its internal reporting and thus is removed from that brand’s aggregate.

It is reasonable that STR has not been able to account for this considering that hotel management teams themselves struggle to keep an accurate assessment of inventory available and the displacement realized throughout the course of relentless projects.

Find Your True North

What metrics are worth obsessing over? CBRE Director of Research Robert Mandelbaum said in the early days of COVID-19, “RevPAR does not tell the full story of what is happening today. The fact is that very few people are traveling, and that message is better depicted via the low occupancy levels.” CBRE is focusing on occupancy rates over revenue. As HotStats wrote in a recent white paper titled The Good, The Bad and The Ugly, “RevPAR has traditionally been the most relied upon metric in the hotel industry. Unfortunately, focusing solely on it is a one-way ticket to financial impairment.” While RevPAR has its flaws, RevPar Index magnifies them exponentially.

Particularly in the COVID and post-COVID environment, RPI and RPI change will be useless for several years. As hotels have closed and adjusted capacity, change metrics will be off for years. Comparing to a baseline of 2019 will also be irrelevant as markets have changed so significantly.

If large brands did not invest in total hotel revenue management and data integration in the boom-years of 2012–2019, there is little reason to expect investment now. Ultimately, owners must be the force for change. Major brands have little incentive to improve the profitability at a unit-level as the contracts are based largely on rooms-sales and RPI. When owners begin awarding contracts to brands based on the entire value of the business, and insist on better measurement of GOPPAR, RevPAR and ROI, then perhaps investment into better metrics will follow.

Break on Through

Dislodging RPI would mean untangling over 50,000 legal agreements between brands, operators and owners. Sometimes, particularly for smaller owners, RPI performance clauses are even in the agreements with lenders. Fighting to remove RPI might be as big of a distraction as paying attention to it in the first place. The winner is not the executive who takes on this thankless fight. The winner is the executive who values their employee’s time enough to stop asking about it.

* All facts shared with permission from All In, Ltd.