2026
From 5 Systems to One Brain: How Urban Leisure Group Unified Their Venue Operations with AI
Urban Leisure Group is a multi-venue hospitality operator in London that was running five separate systems for sales, labour, inventory, events, and accounting. None of them talked to each other. Urban Leisure Group connected Toast, Harri, Marketman, Design My Night, and QuickBooks into a single AI-powered analytics layer. The result: their operations team stopped spending Monday mornings pulling numbers from 5 platforms and started getting automated performance scorecards, venue-level dashboards, and intelligent purchasing suggestions delivered without lifting a finger.
Anubrota Biswas
Growth
About Urban Leisure Group (London Multi-Venue Hospitality Operator)
Urban Leisure Group runs multiple bar and restaurant venues across London. Their business is nights out, great drinks, and the kind of atmosphere you remember the next morning.
Industry: Hospitality, Food & Beverage
Role Leading Adoption: Head of Operations, Finance Director
What They Replaced: Monday-morning spreadsheet assembly from five disconnected platforms
Systems Connected to Milo: Toast (POS), Harri (workforce), Marketman (inventory), Design My Night (events), QuickBooks (accounting)
Access Channels with Milo: Web app + automated Monday email reports
Core Focus Areas:
Weekly venue performance (sales vs labour cost)
Revenue forecasting from historical patterns + live event bookings
AI-powered drink order recommendations
Executive KPI dashboards across all venues
Capital expenditure tracking for new builds
Running a hospitality group means running five businesses at once: retailer, employer, stock manager, events company, and property developer. The data for each lives in a different system. ULG connected all of them and now gets answers in seconds that used to take a morning of spreadsheet work.
Watch the full video of this case study here:
The Challenge
Urban Leisure Group operates multiple venues across London. Their data lives in five systems that were never designed to work together: Toast for sales, Harri for labour, Marketman for inventory, Design My Night for event bookings, and QuickBooks for accounting.
Every Monday, the ops team logged into each platform, pulled numbers by hand, and assembled a performance picture in spreadsheets. It took hours. It was error-prone. And any question that required data from two systems at once (like "what is our labour cost as a percentage of sales this week") meant even more manual work.
They had no data team and no BI platform. Just spreadsheets and a lot of patience.
The Solution
ULG connected all five systems to Milo in under two weeks. No data engineering. No implementation project. The AI understands the relationships between the systems and lets the team ask questions across all of them at once.
What they built:
Automated Weekly Scorecards. Every Monday, leadership gets a full performance report by email. Sales from Toast, labour costs from Harri, period-over-period comparisons, broken down by venue. Zero manual work.
Revenue Forecasting. A dashboard that predicts weekly sales by day of week, using last year's patterns as a baseline and layering on pre-booked event revenue from Design My Night. Managers see what is coming before the week starts.
Drink Order Recommendations. A custom AI app that calculates exactly what to order from each supplier. It takes last week's sales, breaks them into ingredients via Marketman recipes, adds upcoming event demand, and produces a grouped order list. No more guessing quantities.
Executive and Category Dashboards. Top KPIs across all venues (net sales, labour %, GP%, sales per labour hour, waste), plus deep-dives into cocktail and draught beer performance, plus a capex tracker for a new venue build.
The Result
One operations leader started using Milo. Within two months, five more team members joined from finance, purchasing, and venue management. Usage grew over 10x month-over-month. Not because anyone mandated it. Because it actually saved them time.
ULG now has automated reports, forward-looking forecasts, and intelligent ordering. No data team. No BI project. Six weeks from first connection to full operational value.