Access Hospitality has reported a 40% year-on-year increase in the adoption of its AI-assisted menu engineering tools across its UK client base, with the technology moving from a feature that operators explored cautiously to one they are deploying as a standard operational practice.
The data, drawn from approximately 3,200 UK hospitality businesses using the group's EPoS and back-office platforms, shows that operators using AI-driven menu analysis have on average improved gross profit percentage by 3.1 points over a 12-month period — without raising menu prices.
What the tools actually do
Menu engineering as a discipline is not new. The practice of categorising dishes by their contribution margin and popularity — identifying "stars," "plowhorses," "puzzles" and "dogs" in the traditional framework — has been taught in hospitality management courses for decades. What AI-assisted tools change is the frequency and granularity with which the analysis can be performed.
Rather than a quarterly or twice-yearly review of menu performance, the Access Hospitality platform performs continuous analysis across all items, factoring in:
- Contribution margin per dish, updated in real time as ingredient costs change
- Sales velocity by day part, day of week and season
- Table turn impact — dishes that slow service versus those that support faster covers
- Modifier and upsell patterns, identifying where trained prompting by front-of-house teams is converting most effectively
- Weather and local event correlations, flagging when specific categories over- or underperform relative to forecast
The output is a set of recommendations — surfaced in a dashboard reviewed by the operator or their operations team — suggesting which dishes to promote more prominently on the menu, which to remove or reprice, and which represent hidden margin opportunities that are currently underselling.
Operator responses
Access Hospitality shared anonymised case study data with The Mise covering four operators from different sectors.
A 12-site pub group in the Midlands reported that removing five dishes identified as low-contribution, slow-turn items and promoting three underperforming high-margin dishes through repositioning on the menu resulted in a 2.8% improvement in food GP over six months. The operator noted that the same analysis had been done manually 18 months earlier but had taken three weeks; the AI tool produced the same output in under an hour.
A branded casual dining operator with 28 sites used the platform to identify that a specific chicken dish was underperforming in margin terms primarily due to modifier costs — customers were adding components that individually seemed minor but collectively eroded the dish's contribution. A modest repricing of the modifier set recovered the margin without affecting the dish's sales volume.
The GP pressure context
The acceleration in adoption is partly a response to sustained cost pressure. Labour cost increases following the April 2025 National Living Wage uplifts, combined with continued volatility in key commodity categories including oil, poultry and certain produce lines, have left operators looking for margin improvement opportunities that don't require price increases.
"The conversation has shifted from 'this is interesting' to 'show me the ROI and tell me how to get started'," said the group's UK managing director. "Operators are not buying AI. They're buying margin. The AI is the mechanism."
Access Hospitality says it expects adoption of the menu engineering module to reach 70% of eligible clients — those on plans that include the analytics layer — by the end of 2026.