How to Read a Hotel Feasibility Study Like an Owner

A feasibility study is not a prediction. It is an argument, built by a consultant working from assumptions the client, often the owner themselves without quite realizing it, has already signaled they want confirmed. That is not dishonesty; it is how the commission works in practice. The study answers the brief it was given, and the brief usually leans toward yes, because the study is typically commissioned after the owner has already decided to pursue the project. Reading it like an owner means reading past the conclusion to the assumptions that produced it.

We have sat on both sides of these documents, commissioning them and picking them apart before a bank sees them. The pattern repeats across markets and asset types. The math is almost always correct, the multiplication and discounting genuinely sound. The optimism lives upstream of the math, in three or four choices made early in the process that quietly steer every number that follows. An owner who understands where those choices sit can ask five or six pointed questions that change how much a study should be trusted, without needing to rebuild the financial model themselves.

None of this is about distrusting the discipline of feasibility work, which is genuinely useful when done honestly and is a necessary input into any serious financing conversation. It is about reading the document the way a lender's credit committee eventually will, because that reading happens whether the owner does it first, with time to react, or gets surprised by it later, when reacting is expensive.

Comp set selection decides the answer before the model runs

The competitive set chosen for a feasibility study determines the achievable rate and occupancy far more than any subsequent calculation in the model. A comp set built from three aspirational properties the subject hopes to compete with, rather than the properties it will actually compete against on the day it opens, produces a rate assumption the market has not agreed to yet. Ask which properties were excluded and why, not only which were included, since the exclusions usually reveal more about the author's optimism than the inclusions do.

A well-built comp set includes at least one or two properties the owner would rather not be compared to, because that is honesty about where the asset actually sits in its first years of trading, before brand equity and reputation catch up to the physical product. If every comparable in the set is a property the owner admires, the comp set was built backward from the desired answer rather than forward from the market.

It is also worth asking the consultant, before the study is finalized, what would need to change in the comp set for the projected ADR to no longer be defensible, since a comp set that cannot survive one thoughtful substitution was probably too narrow to begin with.

Ramp-up curves are where hope gets smoothed into a model

Every study assumes a ramp-up period from opening to stabilized trading, typically expressed as a percentage of stabilized occupancy reached in year one, two, and three. The convention varies by market and asset type, but the pattern to watch for is a ramp that is faster than comparable openings the owner can independently verify, or a stabilization year that arrives suspiciously close to the end of the projection period so the weakest years carry less weight in a discounted cash flow calculation.

Ask what openings the ramp-up assumption is benchmarked against, by name if the study allows it, or by category and market if confidentiality limits how specific the author can be. A ramp-up curve with no stated benchmark at all, presented as a general convention rather than tied to comparable evidence, is one of the clearer signals that the assumption was chosen to make the model work rather than derived from observed openings.

The macro environment during those comparable openings matters as much as the openings themselves. A ramp-up curve borrowed from properties that opened into a growing market says little about an asset opening into a flatter or more competitive one, and a careful study should say explicitly which conditions its benchmark openings faced rather than presenting a single curve as universally applicable.

Undistributed expenses are the easiest place to under-forecast

Departmental costs, rooms, food and beverage, tend to get reasonable scrutiny because they scale intuitively with revenue and are easy to benchmark against industry conventions. Undistributed expenses, administrative and general, sales and marketing, property operations and maintenance, utilities, are where studies most often understate reality, because they are harder to benchmark against a specific market and easier to assume will behave efficiently under a first-time operator in a location with no trading history to draw on.

A study that shows undistributed expenses as a stable, low percentage of revenue from year one deserves a direct question about which comparable properties support that assumption, at what age and brand tier, and whether the comparables cited were themselves new openings or long-stabilized properties with years of efficiency gains already built in. In NMBR's experience, new hotels rarely run undistributed expenses at stabilized-property ratios in their first eighteen months, regardless of what a clean spreadsheet suggests.

In our experience, sales and marketing in particular tends to be understated for a genuinely new entrant with no brand recognition in its market, since building awareness from zero costs more in the first two years than maintaining awareness for an established property, a distinction base-case models frequently fail to make.

ADR growth assumptions compound faster than owners notice

A study projecting average daily rate growth of a few percentage points annually, compounded over a ten or fifteen year projection period, produces numbers late in the hold period that can look disconnected from the market reality of years one and two. This is not necessarily wrong, rate does grow over a hold period as a market matures, but the compounding effect deserves a sanity check against what the terminal year ADR implies in absolute terms.

Ask whether that terminal figure would still look defensible if the inflation and rate-growth assumptions embedded elsewhere in the same study turn out to be optimistic too, since most studies we read apply a single growth rate consistently across the projection without stress-testing what happens if the underlying market simply grows more slowly than assumed. A number that only survives at the assumed growth rate, with no room for a slower path, is a fragile foundation for a ten year hold.

Watch, too, for a model where operating cost growth is assumed to lag rate growth every year of the projection, since that pairing manufactures an ever-widening GOP margin the market rarely delivers in practice; in our experience costs track revenue more closely than optimistic models assume, particularly on the payroll side.

The sensitivity section tells you what the author actually believes

Most studies include a sensitivity table showing how returns move under different ADR, occupancy, or cost assumptions. This section is frequently the most honest part of the document, because it shows the downside case that the base case narrative is built to avoid emphasizing. An owner reading only the base case narrative and skipping the sensitivity table is reading half the study, and often the more comfortable half.

Ask what happens to debt service coverage, not just to IRR, under the downside scenario, since a lender will ask exactly that question regardless of whether the owner asks it first. A study whose sensitivity table shows debt service coverage falling below a comfortable threshold under a modest, realistic downside case is telling the owner something the base case narrative is not saying out loud.

A sensitivity table that flexes one variable at a time, ADR alone, occupancy alone, cost alone, understates real risk, since downturns rarely arrive as a single clean variable moving in isolation. Ask whether the study includes at least one combined downside scenario, several variables moving against the owner simultaneously, because that combined case is closer to what an actual recession or an oversupplied market looks like.

What owners should take from this

A feasibility study is one input into a larger set of decisions about the project's whole composition, alongside the program, the agreement, and the construction budget, and it earns the same scrutiny an owner would apply to a term sheet, not the passive trust often extended to a technical-looking document.

• Ask which comparable properties were excluded from the comp set and why, not only which were included.
• Benchmark the ramp-up assumption against verifiable comparable openings, not the study's own stated precedent.
• Question undistributed expense ratios directly, especially in year one, against comparable assets of similar age and brand tier.
• Sanity check the terminal year ADR in absolute terms, not only as a growth percentage compounded from a low base.
• Read the sensitivity table before the narrative, since it usually shows what the author actually believes about downside risk.
• Ask what the downside scenario does to debt service coverage, because a lender will ask that question next.