The 18-Month Rule: Why Most Hong Kong Restaurants Fail — And How Survivors Think Differently

The 18-Month Rule Is Real — and It Is Preventable

The failure rate for new food and beverage operations in Hong Kong is not a secret. Industry participants know it. Landlords price it into their lease structures. Suppliers factor it into credit terms. The commonly cited figure — that between 60 and 70% of new Hong Kong restaurants and cafés do not survive their first 18 months of operation — is consistent with the observable reality of any busy commercial street in the city.

What is less understood is why. The conventional explanation attributes failure to generic causes: competition, high rents, economic headwinds. These are not wrong, but they are insufficient. They describe the environment, not the decision. The evidence from operators who succeed and those who fail points consistently to a smaller number of specific, identifiable decision failures — most of which occur before the lease is signed.

This analysis draws on failure pattern data compiled from PLACISE's district intelligence platform, publicly available industry data from the Census and Statistics Department, and structured interviews with F&B operators across Hong Kong's restaurant community.

Fatal Assumption 1: The Wrong District

The most common fatal decision in Hong Kong F&B is choosing a district based on the operator's familiarity with it rather than its suitability for the concept.

This manifests in several recognisable patterns:

The operator opens near where they live. Convenience and familiarity are understandable motivations for a first-time operator, but they are not strategic ones. The residential district that suits the operator's lifestyle may be entirely wrong for their concept's target customer — too affluent, too local, too tourist-dependent, too price-sensitive.

The operator opens in a district they find exciting. The energy of Mong Kok, the prestige of Central, the buzz of Causeway Bay — these are powerful attractors for operators who spend time in these districts personally. But excitement is not demand analysis. A district's street energy is often inversely correlated with its suitability for a new concept: the highest-energy corridors have the most competition, the highest rents, and the least tolerance for a concept that is still finding its feet.

The operator copies a competitor without understanding why the competitor is successful. Observing that a similar concept is doing well in a given district and concluding that the district is therefore suitable for your version of the same concept is a logical error. The successful competitor may have a specific location advantage (ground-floor corner unit, direct MTR exit proximity), a specific customer relationship built over years, or cost economics from their lease vintage that are not replicable at current market rents.

What district selection should actually look like: Effective district selection starts with the target customer — their residential location or daily movement patterns — and works backward to identify which districts offer the highest natural concentration of that customer type. It uses objective data on residential demographics, daytime worker population, visitor flow patterns, and household income to identify district-concept fit. It treats a site visit as validation, not discovery.

C&SD data shows that district-concept mismatch is a contributing factor in an estimated 40–45% of Hong Kong F&B closures within the first 18 months — making it the single largest preventable category.

Fatal Assumption 2: The Wrong Daypart

Hong Kong's F&B market is not one market. It is at least three markets layered on top of each other, segmented by the time of day: the breakfast and early morning market, the lunch market, and the dinner and late-night market. The fundamental economics of each are different, the competitive dynamics are different, and the customer is often entirely different.

Operators who fail to specify which daypart they are primarily serving — and design their concept, staffing model, and economics accordingly — consistently encounter the same problem: revenue that is sufficient during their strong daypart but insufficient to cover total costs when averaged across all operating hours.

The lunch opportunity is systematically underserved in emerging commercial districts. New commercial nodes in Kwun Tong, Wong Chuk Hang, and the emerging Tseung Kwan O corridor have daytime worker populations that have grown faster than F&B supply. A concept that targets the HKD 80–150 lunch set segment in these districts, with tight service and quick turnover, can achieve very strong lunch economics. Many operators in these districts instead build concepts designed around the dinner experience, then discover that evening foot traffic does not materialise at the volume they projected.

The dinner opportunity is systematically overcrowded in prime corridors. In Wan Chai, Causeway Bay, Tsim Sha Tsui, and Central, the competitive density for dinner covers is intense. New entrants without an established following, significant marketing budget, or distinctive concept differentiation will typically run below 60% dinner cover capacity in their first year — a rate that is economically unviable at prime corridor rent levels.

What daypart planning should actually look like: The starting question is not "what hours will we be open?" but "which specific daypart can we profitably dominate?" A concept that owns the lunch segment in an underserved location can achieve viability quickly and grow into other dayparts from a position of cash-flow strength. A concept that disperses its focus across all dayparts in a competitive corridor typically achieves mediocrity in all of them.

Fatal Assumption 3: The Wrong Customer

The third fatal assumption is the most pernicious because it is the least visible until it is too late: operators building a concept for a customer who does not exist at the location they have chosen in the volumes they need.

This occurs in two distinct failure modes:

The tourist-dependent model in a non-tourist location. An operator builds a concept designed to capture tourist spending — Instagram-optimised interiors, international flavours, premium pricing — and locates it in a neighbourhood residential district where the daily customer base is Hong Kong residents who know exactly what they want to pay for a meal. The concept finds a small audience of novelty-seekers and one-time visitors but fails to develop a resident repeat customer base. Without repeat visits, the economics of a single location in a residential district are unworkable.

The resident-dependent model in a high-tourist-dependency location. Inversely, an operator builds a concept designed for locals — authentic, unpretentious, locally priced — and locates it in a tourist corridor where landlords have priced in tourist foot traffic premiums. The concept finds an enthusiastic but insufficient local audience, fails to attract the tourist spend that the rent level requires, and cannot cover costs.

The demographic mismatch. An operator targets a young professional demographic and locates in a district whose residential base skews older or lower-income. They discover, too late, that the density of their target customer at their specific location is insufficient to support their model at their price point.

What customer definition should actually look like: A viable customer definition for a Hong Kong F&B concept specifies: income band, household type, visit frequency motivation (convenience, occasion, destination), and daily movement pattern. That specification is then tested against district-level demographic data before — not after — a lease is signed.

What Survivors Do Differently

The operators who beat the 18-month rule share a common set of behavioural patterns. These are not personality traits or talent differentials. They are systematic habits that can be learned and applied.

They define the customer before they find the location. The founding question is always "who is this for?" — and that question is answered specifically, not generically. Not "young professionals" but "single-income professionals aged 28–38, living in the Kennedy Town–Sai Ying Pun corridor, willing to pay HKD 180–280 for a quality casual dinner within 15 minutes of home."

They run the financials before they fall in love with a unit. The first filter for any specific location is financial viability. Does the achievable rent, at the revenue projection for this concept at this location, produce a positive operating margin? If the answer requires occupancy rates or per-head spends that cannot be justified by the district's available customer base, the unit fails the financial test before it is viewed.

They use data to pressure-test their assumptions. Every material assumption in their business plan — foot traffic, demographic composition, competitor density, achievable rent per comparable unit — is cross-referenced against objective data. Where the data does not support the assumption, they revise the assumption rather than discard the data.

They treat the first 12 months as a learning period, not a ramp-up. Operators who survive build concepts with sufficient financial runway to absorb the inevitable under-performance of the early months. They do not plan to be profitable at month 6 — they plan to understand their customer by month 6, and to be profitable by month 12.

The Role of Intelligence in Decision-Making

The information asymmetry between landlords and new F&B operators in Hong Kong is significant. Landlords know their vacancy history, their renewal rate for comparable tenants, and their district's traffic trends. New operators frequently have none of this information at the point of negotiation.

CBRE's market tracking and PLACISE's district intelligence platform both exist to reduce this asymmetry. The operators who use these resources make materially better-informed decisions — not because intelligence guarantees success, but because it eliminates the most common forms of preventable failure.

The average cost of an F&B setup in Hong Kong — fit-out, equipment, first and last month deposit, working capital float — sits between HKD 1.5 million and HKD 3.5 million depending on concept and size. At that commitment level, the cost of rigorous pre-opening intelligence is trivially small relative to the decision it informs.

Conclusion: The 18-Month Rule Is a Choice

Most Hong Kong restaurant failures are not the result of bad luck, impossible competition, or uniquely difficult market conditions. They are the result of decisions that were made without sufficient information, at the most consequential moment in the business's life.

The operators who beat the 18-month rule do not have better taste, more talent, or a more original concept. They have better information at decision time — and the discipline to let that information override the emotional appeal of a unit they like or a district they are excited about.

The 18-month rule is real. But it is not inevitable.