How Data Turned a Menu Overhaul Into a Revenue Strategy? – Knowledge Center

Property	Details
Property type	Luxury hotel, Asia Pacific
F&B outlets covered	All-day dining restaurant; In-Room Dining; Pool menu
Periods analysed	Lunch, Afternoon Tea, Dinner
Analysis window	Year-to-date, reviewed against trailing 12-week trend
Framework	Signal to Impact

The Signal

The brief was straightforward: the hotel was planning a full overhaul of its all-day dining menu, with In-Room Dining and the pool menu to follow. The stated goals were to reduce operational pressure in the kitchen, improve guest satisfaction, and lift spend per head, particularly at dinner and afternoon tea.

Before the new menu was written, the question was put to the data: what does the current menu actually tell us about where the problems are, and what would a better menu need to do differently?

The starting point was a year-to-date sales mix across all three meal periods, food items only with beverages excluded, combined with KPI data covering covers, average check, table utilisation, and day-of-week breakdowns. A follow-up conversation with the Executive Sous Chef added the operational context the data alone could not provide.

This is the same starting point the Signal to Impact framework is built for: a business question that looks like a menu problem on the surface, but that the data reframes before any decisions are made.

The Analysis

Level 1: The numbers behind the brief

The first layer of analysis established where the restaurant actually stood across its three meal periods.

Period	Revenue (YTD)	Covers	Food check per head	Table utilisation
Lunch	$256K	6,596	$36	62.3%
Afternoon Tea	$128K	4,198	$25	75.2%
Dinner	$289K	9,307	$25	74.3%

Two things stood out immediately. First, dinner was the highest-revenue period with the highest table utilisation, but its food check per head ($25) was below lunch ($36) and equal to afternoon tea. Second, the restaurant was not going to grow its way to more revenue: dinner was already at 74.3% table utilisation and afternoon tea was approaching capacity at 75.2%. The only realistic lever was spend per head.

Level 2: The concentration problem

The second layer looked at how sales were distributed across the menu. In every meal period, a Pareto pattern emerged: roughly half of all food orders were coming from just 10 dishes, out of 60 to 80+ active food SKUs.

Period	Active food SKUs	Top 10 items: % of volume	Top 10 items: % of revenue
Lunch	81	52.2%	62.8%
Afternoon Tea	61	53.4%	53.6%
Dinner	69	51.2%	59.6%

The kitchen was operating with the complexity of a 70 to 80 item menu to generate sales concentrated in 10 dishes. With four ranges and 1.5 to 2 chefs per section at peak, the same brigade covering dinner, In-Room Dining, the pool, and events simultaneously, the operational cost of that long tail was significant and largely invisible in the revenue numbers.

The conversation with the kitchen team confirmed it. The pressure was not coming from the high-volume dishes. It was coming from low-frequency items requiring standing mise en place, specific ingredients, and preparation time that pulled the kitchen away from the items guests were actually ordering most.

Level 3: Volume versus revenue, the dinner gap

The third layer compared the volume ranking and the revenue ranking at dinner. They told very different stories.

Item	Dinner orders (YTD)	Revenue	Avg. price per order
Signature noodle dish	347	$7,000	$20
Side of fries	320	$2,000	$6
Sharing snack	233	$2,800	$12
Classic sandwich	197	$4,400	$22
Premium beef main	164	$8,500	$52
Premium seafood rice	159	$6,500	$41
Premium fish main	84	$4,200	$50

A side of fries was the second most ordered food item at dinner, 320 orders at an average of $6. In the same period, the three premium mains together represented 407 orders at an average of $48. The kitchen was handling comparable order volumes for a revenue ratio of roughly 8:1 in favour of the premium items.

Premium dishes accounted for only 5.8% of dinner covers combined. At an average of $45 per order, these were the items the menu needed guests to reach for more often. The data was not a demand problem. It was a guidance problem: the menu and the service were not directing guests toward the items that would move the check.

The conversation with the kitchen team added an important qualification. The low premium penetration was partly about price perception: guests had given feedback that certain items felt expensive relative to their portion. Steps had already been taken to adjust. But the team also acknowledged that staff were not consistently confident recommending premium items, and that the previous menu had positioned the restaurant closer to a cafe format than a luxury dining outlet. The brand perception gap was real, and the menu alone could not close it.

Level 4: The quality signal inside the volume numbers

Several of the highest-volume items at dinner had active guest complaints attached to them. A seafood sandwich, sixth highest revenue at dinner, had recurring feedback about bread quality. A pasta dish was in the top four by revenue but carried a trail of negative guest responses. A popular local soup had consistent complaint patterns despite 135 dinner orders in the period.

This mattered for a specific reason. Removing or fixing a low-volume complaint dish is operationally manageable. Fixing a complaint dish that is also in the top six by revenue is urgent: it is both a quality risk and a revenue risk. Letting it continue into a major hotel inspection without resolution would be the worst of both outcomes.

The Insight

The menu overhaul was not one problem. It was at least four:

A complexity burden: the kitchen was running 70 to 80 food SKUs per period with half the revenue coming from 10. Every additional SKU below that core had an operational cost that was not visible in the check average but was being paid in kitchen pressure, prep time, and consistency risk.
A guidance failure: premium items existed on the menu and guests were ordering them, but at a penetration rate of 5.8% of covers. The menu format, service confidence, and brand positioning were all working against guests reaching for the higher-value options.
A quality conflict: the dishes with the most guest complaints were not marginal items. They were revenue anchors. Fixing them was not a later phase task. It was pre-launch.
A pricing perception gap: at least one premium item had received enough feedback about value relative to price that portions had already been adjusted. The data could not resolve whether the issue was genuinely about price, about presentation, or about the brand not yet supporting the price point. That was a question for the operating team, but it was a question worth asking explicitly before setting new menu prices.

What the data reframed, above all else, was the target. The conversation had started with a menu change. The data made clear it was a spend-per-head problem: dinner food check at $25 per head, with a target of reaching $32 to $34, a 35% increase, not through more guests, but through the same guests ordering differently.

Where This Goes Next

The three-phase menu plan the team had developed was reviewed against the analysis above. The direction was sound: the right items were being kept, the right items were being removed, and the SKU reduction was significant. The questions the data raised were about execution, not concept.

Open question	Status
Two high-revenue dishes with active quality complaints: resolve before Phase 1 launch during the hotel inspection window.	Pre-launch action
Front of house training on the four Phase 1 mains: narrowest menu, best window to build staff confidence.	Phase 1 action
Premium fries upgrade offer: over 300 basic fries orders YTD at $6 each. Script the upgrade to the premium version at $15 for every order.	Phase 1 action
Two popular mid-tier items removed with no direct Phase 1 replacement, representing $6,200 in combined dinner revenue. Monitor the gap.	Watch Phase 1
Premium sharing starter introduced in Phase 2: highest-value new item across all three phases. Train front of house to lead with it as an aperitif and pair with the beverage programme.	Phase 2 action
Two new lower-priced mains added in Phases 2 and 3: necessary for menu balance but lowest-priced options. Agree a front of house approach to avoid them becoming the default order.	Phase 2 to 3 action
Signature noodle dish price reduction planned for Phase 3: highest-volume dinner item at 347 orders YTD. Confirm the reduction is intentional given the annual revenue impact.	Pre-Phase 3 confirmation
Premium item penetration: from 5.8% of covers (1 in 17 guests) toward 10 to 12% (1 in 9). This is the single metric that will determine whether the spend-per-head target is reached.	Ongoing, measure monthly

Key Lessons for F&B Leaders

A menu complexity problem and a spend-per-head problem are not the same problem: even when they show up together. Fixing the first without addressing the second gets you a simpler menu at the same check average.
Volume and revenue rankings tell different stories, and both matter: The volume list shows where the kitchen pressure is. The revenue list shows what the check is actually built on. An item can be a kitchen burden, a revenue anchor, and a quality risk at the same time, and each of those facts calls for a different response.
Complaint dishes that are also revenue anchors are the highest-priority fixes: not the lowest. The instinct is often to focus on improving the dishes guests love. The data says fix the dishes guests keep ordering despite not loving them first.
Staff confidence is a revenue variable, not a training nicety: If the team is not comfortable recommending premium items, the menu price architecture is irrelevant. The window when a menu is at its leanest, fewest dishes and clearest focus, is the best window to build that confidence.
The menu creates the conditions, but it does not close the deal: The difference between a 5.8% premium penetration rate and a 10 to 12% rate is not a new menu. It is a team that knows what to say when a guest sits down.

Relevance for Other Properties

This pattern is worth examining at any all-day dining outlet where:

Table utilisation is high but spend per head has not kept pace: the constraint is not footfall, it is conversion.
The same dishes dominate every meal period: the menu may be functioning as a single all-day cafe rather than three distinct experiences.
Premium items are present but infrequently ordered: the gap is almost always guidance, not demand.
Quality complaints are concentrated in high-volume dishes: these are the items most likely to be ordered again before a fix is in place.
A menu overhaul is being driven by operational pressure alone: the data will almost always surface a revenue story alongside it that changes the design brief.