Restaurant ventures failures: six mechanisms, one named case

A room with a hundred seats cannot serve a hundred and one. A room with a hundred seats and forty guests still pays for a hundred. Every restaurant failure is a version of that sentence, and no famous name changes either half of it.

This page explains restaurant ventures failures as mechanisms: six ways a room dies, and what a famous name adds to each. It then tests them against one named public record, Planet Hollywood, the chain launched with Hollywood stars as backers that ended up in bankruptcy.

The smaller case pattern further down is a composite rather than a named room, because attributing a single closure to one person's judgment requires knowing things that never leave the operating company.

What to take away

The ceiling and the floor

The ceiling is physical. Seats, multiplied by how many times each is used in a service, multiplied by what each guest spends. You can raise spend a little with the menu and turns a little with the booking system. You cannot add a floor to the building.

The floor is contractual. Rent for the term, minimum staffing to open the doors at all, utilities, the loan on the fit-out, insurance, and the conditions on a permit issued locally under an adopted version of the FDA Food Code. It is paid whether the room is full or empty.

Between them is the margin at capacity, and the question every venue answers in its first year is what fraction of capacity ordinary demand delivers. The restaurant ventures page sets out the arithmetic; this page is about the ways the answer turns out to be too low.

Landlords and lenders commonly look for rent at 6 to 10 percent of sales. A site that needs 12 percent costs more than the trade around it supports.

Six ways a room dies

The trough

The novelty period ends and the room finds out what it earns from people who come back. If that figure clears the floor, nothing happens. If it does not, the room has three moves: change the offer, renegotiate the rent, or close.

Run the test on one ordinary day. A 100-seat room at month 18, on a Tuesday: 58 covers at a $55 check is $3,190 at dinner, plus about $1,800 at lunch, so the day lands near $5,000.

Rent of $14,000 a month is about $460 a day, so that Tuesday clears rent and part of the day's payroll. Across a month the same day repeats into about $150,000 in sales, and rent sits at 9 percent, inside the typical 6 to 10 percent band.

Let the same Tuesday fall to $4,300, which is what a trough looks like, and rent is 10.8 percent. The lease did not change; the room left the band.

The trophy build

A well-known name attracts a landlord with a better building and a designer with a bigger budget. Both raise the floor, often on a site chosen for prestige rather than footfall. The build looked like an advantage and was a commitment.

A full-service build commonly runs $150 to $400 per square foot, so a 100-seat room at about 15 square feet a seat lands around $225,000 to $600,000. A build that only works on launch demand is a trophy build.

The key-person departure

Some rooms are a person. When that person leaves, whether the chef, the operator or the host, the concept goes with them and the lease stays. The name on the awning may never have been the person the room depended on.

Here is the pattern without names. A 90-seat chef-driven room opens with a chef who owns a small stake and whose cooking is why people book. In month three the chef leaves for a larger group.

The brigade stays, the menu stays, the room keeps trading. Repeat visits fall anyway. The license moves to a new officer, the lease runs another 14 months, and the entity closes.

The landlord's move

Leases have reviews and break clauses, and landlords have plans. A room can be trading acceptably and still lose its site because the review priced it out or the building is being redeveloped. From outside, this is indistinguishable from failure. From the lease, it is a landlord exercising a right.

Rent reviews commonly fall at year 5 and every 5 years after, with a break clause usually at year 5 or year 10.

The cash-cycle illusion

Guests pay tonight; suppliers are paid later. For a while, the bank balance rises even in a room losing money on each service. When the supplier terms catch up, the loss arrives all at once. The celebrity companies by industry page notes this is specific to businesses that collect instantly and pay on terms.

Supplier terms commonly run 30 days. A supplier who shortens to 14 days flips the balance within one payment cycle.

The group decision

Multi-room operators run each venue as its own entity so that they can close one without the others. When they do, the closed room is described as a failure and the decision was a portfolio one: the site with the worst margin was let go so that the group's guarantees were not called.

The paperwork is the mechanism: separate entities, separate bank accounts, no cross-guarantee between rooms.

A named case: Planet Hollywood

Planet Hollywood is the best-documented star-backed restaurant failure. It was launched in New York City on October 22, 1991, at the base of two office buildings at 130 and 140 West 57th Street, with the backing of Sylvester Stallone, Bruce Willis, Demi Moore and Arnold Schwarzenegger. The company was founded by Robert Earl, and the recruited actors were paid for their appearances and endorsements through an employee stock ownership plan.

The Planet Hollywood company history shows how little of the venture the names owned. As of the May 1996 IPO circular, the celebrities were only minor holders of stock, with Earl and Barish owning 57.2 percent and a 24.2 percent stake owned by Ong Beng Seng. The company went public on the New York Stock Exchange in April 1996 at $18 a share; by 1999 the shares were below $1. It recorded a $40 million loss in the fourth quarter of 1997 and $244 million in losses in 1998, and the same history describes a core chain neglected during diversification, with mediocre food, high menu prices and below average service. At its 1990s peak it operated more than 60 locations worldwide; it filed for bankruptcy on October 12, 1999, and the restructuring left 35 locations in 2000.

Read against the six mechanisms, the record fits the trophy build and the trough more than anything the stars did. The names filled each new room at its opening gala; ordinary repeat demand, at those prices and that service, did not hold up the floor of a rapidly expanding chain. And the question in the last FAQ below has a documented answer here: the famous backers were minor shareholders, and the controlling stakes sat with Earl, Barish and Ong Beng Seng.

What a name adds to each

A name changes the cost of the decision more than the arithmetic of the room. A closure with a recognizable name attached is news, and news delays the decision past the cheapest moment to make it. The celebrity brands page explains why the same pattern appears wherever attention is front-loaded.

What the record shows

The license register shows who operated the room and when they stopped. The company register, whose filing obligations are described in the guide to registering a business, shows when the entity was dissolved. The court docket shows whether suppliers or the landlord sued. The planning register shows the build.

None of them show the rent, the guarantee, the trough, or which of the six mechanisms applied. A defensible account of a closure, following the portfolio case studies method, states the dated facts and leaves the mechanism as an open question. Anything more confident is a guess about a private contract.

No register records why a room stopped trading, so a figure for landlord-driven closures would be invented.

What the registers do give is a lag. An entity that stops filing is administratively dissolved on the state's annual report cycle, commonly 12 to 24 months after the last license renewal.

First-time actions

Four decisions, all made before the lease is signed.

Common questions

Can a restaurant that is full every night still fail?

Yes, if the floor is above what a full room earns. A trophy rent on a room with modest spend per guest does this. Full is not the same as profitable.

Why can't a popular room just expand?

Expansion means a second lease, a second fit-out, and a second trough, in a different building with different footfall. It is a second venture, not a scaling of the first.

Is closure always the operator's decision?

No. Landlords, lenders and guarantors all have rights that can end a room. Franchised sites add the franchisor; the franchises and licensing page explains what that agreement can do.

Does a closed venue mean the named person lost money?

Only if they held equity in the venue entity or guaranteed its lease. Under a license, they lost future royalties. The record does not say which, and coverage usually assumes the wrong one.