Compare
Period
Impossible MCR (7d) vs Church LDS (prev 7d)
Group Turnover · Today
£487,294
↑ +12.4% vs. same Sat LW
Budget £442,000 · Variance +£45,294
Above Break-Even
54/68
↑ +7 vs. yesterday
9 tracking · 5 below · 0 closed
Gross Margin · Live
71.2%
↑ +0.8pp vs. budget
GP £346,953 · Target 70.4%
Labour · Rolling 7d
23.4%
↓ -0.6pp vs. plan
£113k today · -£2.9k
Cash · Consolidated
£2.41M
34 days cover
Sage sync · 00:03 ago
♦ LIVE
01
Global position
Global Footprint
68 venues, 4 continents
Above Break-Even 54
Tracking 9
Below 5
Closed 0
Saturday snapshot · 23:47 GMT
Live Rankings
Venue leaderboard
| Venue | Today | vs LW | Trend | Break-Even | GP % | Status |
|---|
Group Rollup
Hour by hour
Peak (so far)
22:00–23:00
£89,420 · +18%
Forecast Close
£648,200
+£82k on budget
Pace vs Forecast
+4.7%
Updated 23:47
GeographyBy region
United Kingdom
49 venues · 44 trading
£381.4k
BE: 36/41 · 87.8%↑ 11.2%
Balearics & Med
8 venues · 8 trading
€58.2k
BE: 5/5 · 100%↑ 22.1%
USA
8 venues · 8 trading
$42.1k
BE: 4/6 · 66.7%– Tracking
Middle East
3 venues · 3 trading
62.3k AED
BE: 2/2 · 100%↑ 8.4%
Croatia
1 venue · 1 trading
€6.8k
BE: 1/1 · 100%↑ 4.1%
Festivals & Events
3 active · Lost Village
£12.4k
Pre-sale · Ticketing– Off-peak
Live LedgerP&L snapshot
Line
Today
WTD
MTD
vs Budget
Net Sales
£487,294
£2.84M
£11.2M
+£487k · +4.5%
Wet Sales (68%)
£331,359
£1.93M
£7.62M
+3.2%
Dry Sales (22%)
£107,205
£624k
£2.46M
+8.1%
Tickets (10%)
£48,730
£284k
£1.12M
+6.7%
COGS
£140,341
£820k
£3.25M
-0.4pp
Gross Profit
£346,953
£2.02M
£7.97M
+£58k
Direct Labour
£113,827
£664k
£2.62M
-£2.9k
Door & Security
£28,700
£167k
£661k
+£1.2k
DJ & Entertainment
£42,100
£245k
£967k
on plan
Other Variable
£19,480
£113k
£447k
-£0.5k
Contribution
£142,846
£831k
£3.28M
+£53k · +18.9%
Fixed Overhead (daily)
£86,400
£605k
£2.41M
fixed
EBITDA · today
£56,446
£226k
£869k
+£48k YTD
Product MixLive categories
Spirits
£171,204
35.1%
Cocktails
£101,328
20.8%
Beer / Lager
£58,475
12.0%
Entry / Tickets
£48,730
10.0%
Food
£42,130
8.6%
Wine / Champagne
£36,420
7.5%
Soft Drinks
£22,340
4.6%
Exceptions · Auto-TriagedPriority alerts
Tower Ballroom, Hull — below break-even
Pace 47% of BE by 23:30 · Forecast £8.2k vs £14.8k BE · GM recommends last-hour drinks promo
23:41
Church, Leeds — labour overrun
Live labour 31.4% vs target 24% · 2 staff off-rota · Contact DM to stand down early
23:28
Digital Newcastle — stock variance
Vodka Absolut: 4.2L POS vs 6.1L dispensed · +45% variance · Flag for AM review
23:14
Impossible Manchester — all-time Sat record
£38,420 by 23:30 · Prior record £37,100 (15 Feb) · Surpassed with 2.5h to trade
23:30
Klute, Durham — till offline 6 min
Zonal terminal #3 disconnected 23:42–23:48 · Est. £180 unrecorded · Auto-reconciliation pending
23:48
AMOK Mallorca — Cirque takeover sold out
Capacity 100% · +€4.2k upsell on tables · Wait list: 247 guests
22:58
Sage Line 50 — EoD posting ready
61 venues reconciled · 5 pending close · Journal batch #TI-2026-04-18 awaiting approval
23:00
♦ Radar · Saturday Forecast Engine
Which Saturdays print money
Next 12 Saturdays
68 venues · 7 regions
97% accuracy
Group Business Overview · Selected Saturday
Revenue by Region · Selected Saturday
Why This Saturday · Key Factors
Regional Heat · Selected Saturday
Top 5 Venues · Selected Saturday
02
The estate
Region
Status
Sort
03
The numbers
Group MTD · Revenue
£11.2M
↑ +8.4% vs. April LYBudget £10.4M · +£812k
Group MTD · EBITDA
£869k
↑ +£48k vs. plan7.8% margin · target 7.2%
Cash at Bank
£2.41M
34 days coverPeak £3.1M · Low £1.8M
Unreconciled (Sage)
£3.2k
99.97% auto-match5 venues pending close
Month to DateFrom revenue to EBITDA
Daily Closing BalanceCash position · 30 days
30-day Peak
£3.14M
Sat 12 Apr · post-wk takings
30-day Low
£1.82M
Tue 1 Apr · post-payroll
Avg Daily Burn
£71k
rolling 30 days
Ranked by MTD EBITDA £Top earners
| Venue | Revenue | EBITDA | Margin |
|---|
Ledger HygieneSage reconciliation
Auto-reconciled (24h)
99.97%
2,847 of 2,848 transactions
Manual touches saved
47h/wk
vs. pre-Pulse baseline
Journal batch TI-2026-04-17 posted
61 venues · 2,847 txns · £432,180 revenue · Posted 05:03 GMT
05:03
1 unmatched transaction · Klute Durham
Card auth £142.40 · No POS line · Terminal offline window · Review by FD
04:47
VAT Return Q1 2026 · filed
£394,220 · HMRC acknowledged · Source data 100% machine-extracted
Apr 07
04
The people cost
Labour £ · Tonight
£113,827
↓ -£2.9k vs. plan412 on shift · £276 avg/head
Labour % · Live
23.4%
↓ -0.6pp vs. 24.0%Range 18–34% by venue
Productivity · £/staff-hr
£142.80
↑ +£8.20 vs. LWBenchmark £118
Flags
3
Church · Hull · YorkOver-rota detected
Rolling 24 Hours · All 68 VenuesLabour heatmap
05
The pour
Group Shrinkage · 7d
2.8%
↓ -1.4pp vs. pre-PulseIndustry benchmark 3–8%
£ Saved · 30d
£42k
↑ vs. baselineAnnualised £512k
Variance Flags
14
across 6 venues3 critical · 11 watch
Pour vs POS
97.2%
↑ +0.4pp vs. LW24h rolling match rate
Live Exceptions · by SKUVariance detector
SKU / Category
POS (L/units)
Dispensed
Variance
£ Exposure
Flag
06
The event economics
Marketing ROI · 30d
4.8x
↑ +0.4x vs. LMSpend £94k · Return £451k
Tickets Sold · Tonight
14,240
↑ +18% vs. LW£203k entry revenue
Guest List Conv.
68%
↑ +4pp4,820 RSVP · 3,278 arrived
Avg Spend / Head
£34.20
↑ +£2.80Peak venue: £86 (One 194)
06·a
Tonight’s flagship events
♦ AI Revenue Engine · Next Saturday · 25 April 2026
Group forecast: £512–£548k
ML model predicts £528k central case. Baseline (zero marketing) = £468k. Marketing uplift target = £60k via autonomous budget allocation across 68 venues. Model accuracy: 96.9% over last 13 weeks.
8-Week RollingSaturday actual vs forecast · baseline + marketing uplift
07·a
How the prediction engine works
1
Ingest
EPOS data from 68 venues · 412 terminals · Real-time
2
Enrich
32 external factors per venue per day · Weather, events, macro
3
Predict
ML model · Baseline revenue per venue · 15-day window · 97% accuracy
4
Attribute
Actual − Baseline = Marketing uplift · Confirmed ROAS per channel
5
Optimise
AI agents auto-allocate marketing spend · Per venue, per channel · Autonomous
07·b
70+ external factors tracked daily
Weather & Environment · 8 factors
Temperature
Hourly · per venue postcode · Met Office
+2.1%
16°C Sat 25 Apr. Beer garden threshold >14°C met. Manchester +2.8%, Newcastle +1.4%. Every °C above 14 = +0.8% footfall. Below 8°C = indoor-only model activates.
Rainfall & Precipitation
Hourly probability · 15-day lookahead
+1.8%
0% rain Sat 18:00–04:00. Dry Saturdays = +6–8% footfall vs rainy. Rain before 18:00 has 0.4x impact of rain after 18:00. Heavy rain (>4mm/hr) = −12% avg.
Wind Speed & Gusts
Beaufort scale · outdoor seating threshold
0%
8 mph (light). Below 15 mph threshold — no terrace impact. >25 mph closes outdoor areas at Impossible MCR, Gorilla, Hangar 8289. Storm warnings trigger rota reduction.
Sunset & Daylight Hours
Golden hour timing · terrace economics
+0.6%
Sunset 20:22 BST. Late sunset = extended terrace covers. Impossible MCR terrace rev +£3.2k on evenings with sunset after 20:00. Ibiza: sunset 20:40 — pre-club dining peak.
UV Index & Sunshine Hours
Beach & pool venue driver · Ibiza/Dubai/Hvar
+3.4%
UV 6 in Ibiza (high). Correlation: high UV days = +14% pool party rev at Hangar 8289. Dubai >UV 10 = indoor migration — mall venue uplift. UK venues unaffected.
Humidity & Feels-Like
Comfort index · outdoor willingness
0%
62% humidity, feels like 15°C. Dubai: >80% humidity = 0% outdoor trade even at 28°C. UK rarely a factor. Ibiza >85% humidity + >30°C = shorter outdoor sessions.
5-Day Weather Forecast
Forward-looking · rota & marketing planning
+1.2%
Next 5 days: dry until Wed, rain Thu, dry Fri/Sat. Model pre-books extra terrace staff for Sat. Marketing engine creates “sunshine Saturday” creative 72hrs ahead. Forecast accuracy: 91% at 5 days.
Severe Weather Alerts
Met Office amber/red · Storm names
−0%
None active. Storm Ewan (Feb) caused −38% across Northern venues on one Saturday. Model now auto-scales marketing down 48hrs before amber warnings. Saves £4–8k wasted ad spend per event.
Students & Education · 10 factors
University Term Dates
Durham · Leeds · Newcastle · Manchester · York · Hull
−2.1%
Durham Easter term ends 13 Jun — Klute footfall drops 62% within 48hrs. Newcastle stable to 23 May. Manchester to 6 Jun. York to 20 Jun. Hull to 30 May. Each uni departure = −200–400 weekly covers at nearest venue.
School Half Terms
England & Wales · 6 regions tracked
0%
Next: Mon 26 May – Fri 30 May (most regions). Effect: family pubs/restaurants +12–18%, nightclubs −4% (parents stay home). Ibiza flights spike +22% during half term. Manchester school hols differ from Newcastle by ±1 week — staggered impact.
Easter Holidays
Uni + school overlap · 2–4 week window
−1.8%
Easter hols ended 13 Apr. Student venues (Klute, Church, Fibbers) saw −28% during break. Non-student venues (Impossible, Digital) unaffected. Easter is the longest uni break — model weights it 2x vs half term.
Freshers’ Week
Sep/Oct · biggest student event of the year
+42%
Sep 2025: Klute +68%, Church +54%, Digital +38%. Newcastle Freshers = 8,200 new students. Durham = 4,400. Leeds = 7,800. Model pre-allocates £28k marketing budget 6 weeks ahead. Single biggest revenue week for 12 venues.
Student Loan Drop Dates
SLC maintenance payments · 3 per year
0%
Drops: Sep (largest — £2,500 avg), Jan (£2,200), Apr (£2,200). Apr drop was 2 Apr — spike visible w/c 7 Apr at Klute (+34%), Church (+28%), Fibbers (+22%). Effect lasts 10–14 days then normalises. Currently in decay phase.
Exam Season
Jan + May/Jun · attendance crater
−3.4%
Summer exams start: Durham 12 May, Newcastle 19 May, Manchester 20 May, Leeds 13 May. Klute drops −45% during Durham exams. Church −32% during Leeds exams. Model pre-adjusts baseline & cuts student-targeted marketing to zero. Redirects budget to non-student venues.
Graduation Weeks
Jun/Jul · family spend premium
+8.2%
Durham graduation: 24–28 Jun. Newcastle: 7–11 Jul. Manchester: 14–18 Jul. Family dining spend 2.4x avg student spend. Impossible MCR graduation week = £52k (vs £38k normal). Premium wine/champagne uplift +180%. Model switches to premium-skew pricing.
Summer Ball Season
May/Jun · pre-party & after-party trade
+2.8%
Newcastle Summer Ball: 14 Jun (2,400 tickets). After-party at Digital = +£8.2k. Durham Castle Ball: 21 Jun — Klute after-party sells out. Leeds Ball: 7 Jun — Church benefits. Pre-party bookings trackable via Fatsoma/FIXR APIs.
Reading Week / Study Leave
Mid-term · varies by institution
−1.2%
Durham reading week w/c 28 Apr. 40% of students leave the city. Klute −22% that week. Newcastle reading week w/c 5 May. Some students go home, some stay — net impact lighter (−8%). Not all unis have them — tracked individually.
A-Level Results Day
Mid-Aug · 18th birthday cohort
+6.4%
Thu 13 Aug 2026. Biggest under-21 night of the year. Digital NCL +£14k, Church LDS +£8k. Strict ID checking — extra door staff pre-booked. Model creates targeted campaign 7 days ahead. Results day Thursday → massive Friday/Saturday.
Flights & Tourism · 12 factors
Outbound Flights — MCR
Manchester Airport · 18–34 pax drain
−4.2%
6 MCR→IBZ flights today (Ryanair, easyJet, Jet2). 1,240 pax, 72% aged 18–34. Revenue drain: −£42.2k from Manchester venue catchment. Also 3 MCR→Mallorca (380 pax), 2 MCR→Hvar via Split (180 pax). Total 1,800 target demo leaving the city.
Outbound Flights — NCL
Newcastle Airport · Jet2/Ryanair routes
−2.8%
3 NCL→IBZ flights (620 pax, 68% 18–34). 2 NCL→Alicante (440 pax, older demo — lower impact). 1 NCL→Mallorca (180 pax). Drain on Digital, Tup Tup, Neptune. Heaviest impact on Friday evening departures — lose pre-loading crowd.
Outbound Flights — LDS/LBA
Leeds Bradford Airport · limited routes
−1.6%
2 LBA→Mallorca (Jet2, TUI — 380 pax, 64% 18–34). 1 LBA→Faro (220 pax). Smaller airport = less drain than MCR/NCL. Church and Get Baked see −8% on peak departure days. Friday evening flights are worst — lose the whole weekend cohort.
Outbound Flights — LHR/LGW/STN
London airports · massive volume, diluted impact
−0.8%
14 LHR→DXB (3,800 pax, only 38% 18–34). 22 LHR/LGW→IBZ (2,200 pax). London catchment so large (9M+) that 6,000 departures = <0.1% of going-out population. Impact on Emerald Theatre negligible. Dubai venues gain.
Inbound Arrivals — Ibiza
IBZ airport · revenue gain at destination venues
+8.6%
42,000 arrivals this month. 64% aged 18–34. 41% British (our core demo). Hangar 8289 tonight: 380 UK arrivals within 4km. Avg spend €85/head vs €52 locals. British tourist premium = +€33/head. Payday weekend amplifies — UK tourists spend 18% more post-payday.
FOMO Bounce Effect
Post-holiday Instagram · going-out intent uplift
+2.5%
High Ibiza Instagram volume this week from MCR/NCL users. Followers who didn’t go see Stories → fear of missing out → +5–8% going-out intent following weekend. Measurable via Google Trends “bars near me Manchester” uplift. Effect peaks 5–7 days post-holiday and lasts ~10 days.
Flight Pricing & Seat Load
Skyscanner/Google Flights · demand proxy
+0.4%
MCR→IBZ avg £89 return (cheap). Seat load factor: 94% (near full). When flights sell out, drain peaks. When prices >£200, 30% of 18–24s stay home — model detects price elasticity and adjusts Manchester baseline up. Ryanair flash sales trigger −spike alerts 48hrs ahead.
New Route Announcements
Airline schedule changes · OAG data
+0.2%
Jet2 added 2x weekly NCL→Hvar (Jun–Sep 2026). 180 seats/flight = 1,440 extra departures/month from Newcastle catchment. Model updates long-range drain forecast. Also: easyJet expanding LBA→Dubrovnik — indirect Hvar feeder via bus/ferry.
Cruise Ship Arrivals
Ibiza · Mallorca · Hvar · Dubai
+1.2%
MSC Virtuosa docking Ibiza Tue 22 Apr (3,400 pax). 12% aged 18–34. Crew shore leave adds ~200 nightlife visitors. Hvar: yacht arrivals 148/wk — premium spend €120/head. Dubai: QE2 events drive footfall to DIFC. Tracked via MarineTraffic AIS data.
Visa & Entry Changes
ETIAS · UAE visa-free · policy shifts
0%
ETIAS (EU entry fee) delayed again — no current impact. UAE visa-free for UK nationals = zero friction. If ETIAS launches (€7 fee), minimal impact on leisure travel. If Schengen tightens, Ibiza bookings could shift −2–4% to non-EU destinations (Dubai, USA).
Google Trends: “Flights to”
Forward-looking demand proxy · 4–8 week lead
+1.8%
“Flights to Ibiza from Manchester” search volume +34% vs 4 weeks ago. Correlates 0.82 with actual pax 4–8 weeks later. Free proxy for ForwardKeys (£50k/yr). Also tracking: “Ibiza clubs 2026”, “best nightlife Ibiza”, “pool parties Ibiza”.
Airline Capacity Changes
Seat count increases/decreases · seasonal schedules
+0.6%
Ryanair +30% MCR→IBZ capacity for Jun (vs Jun 2025). easyJet maintaining LGW→IBZ at 2x daily. Jet2 added A321neo on NCL→IBZ (220 seats vs 189). More seats = more drain from UK, more gain in Ibiza. Model adjusts both origin & destination forecasts.
Economic & Macro · 12 factors
BoE Interest Rate
Base rate · Disposable income lag
+0.8%
4.25% (cut Feb −0.25pp). Next MPC: 8 May — market pricing 60% chance of cut. Mortgage holders have ~£30/mo extra. Effect visible 30–60 days after cut. 2.2M tracker mortgage holders benefit immediately. Fixed rate renewals = delayed wave.
CPI Inflation
Consumer price index · spending confidence
+0.6%
2.8% and falling (from 3.4% peak). Target: 2.0%. Every 0.5pp drop = +£8–12 per head per month discretionary spend freed up. Nightlife elasticity: 1.4x — hospitality benefits more than retail from falling inflation. ONS release: 16th of each month.
Real Wage Growth
ONS data · wages vs inflation
+1.2%
Positive for first time in 18 months. Nominal wages +4.1%, CPI 2.8% = real +1.3%. Hospitality sector wages growing +3.2% (NLW increase Apr). Regional: NE +2.8%, NW +3.4%, Yorkshire +2.6%. Higher wages = higher going-out frequency. Lag: 4–6 weeks from paycheque to behaviour change.
Employment / Unemployment
Regional ONS · claimant count
−0.3%
UK 4.1%. Hull 5.8% (drag on Tower). Manchester 3.8% (strong). Newcastle 4.2%. Leeds 3.9%. Durham 3.4%. Youth (18–24) unemployment: 12.4% nationally. Every +1pp youth unemployment = −2.2% nightclub revenue in that city. Hull youth: 16.1% — structural problem.
Consumer Confidence
GfK index · spending intent
+0.4%
GfK +2 — best since Sep 2024. Major purchase intent sub-index +5. Going-out intent correlates 0.72 with GfK. Every +1 GfK point = +£1.80 avg spend per head. Monthly release (last Friday). Model ingests within 2 hours of publication.
Payday Cycle
Monthly + 4-weekly patterns · per employer type
+3.2%
25 Apr = most common UK payroll date (last Friday). NHS pays 27th, civil service 25th, retail weekly. Avg spend +£4–6 per head on payday weekends. Northern venues see stronger effect (+4.2%) than London (+1.8%). 4-weekly pay (Amazon, warehouse) creates off-cycle mini-spikes.
FX Rates
GBP/EUR · GBP/USD · GBP/AED
+0.2%
GBP/EUR 1.172, GBP/USD 1.298, GBP/AED 4.77. Strong pound helps consolidated P&L for Ibiza/Dubai/USA. Also: weak GBP makes Ibiza feel expensive to UK tourists — reduces drain from UK cities. Every −0.05 GBP/EUR = +0.8% MCR/NCL retention.
Fuel & Energy Prices
Petrol · Ofgem cap · operating costs
+0.1%
Petrol £1.38/L (stable). Ofgem cap down 8% from Apr. Operating cost saving £800–1,200/venue/month feeds EBITDA. Taxi fares track fuel — cheaper Ubers = easier late-night transport = +0.4% footfall in suburban venues. Dual impact: margin + demand.
Retail Sales (ONS)
Leading indicator · consumer spending proxy
+0.3%
Mar retail sales +1.2% MoM. Strong retail = confidence proxy. Correlation with hospitality spend: 0.58 (moderate). Clothing/footwear sub-index more relevant — people buying outfits = going out. ONS release 3rd Friday of month. 2–4 week leading indicator.
Housing Market
Rightmove · mortgage approvals · wealth effect
+0.1%
House prices +2.4% YoY (Rightmove). Wealth effect: homeowners feel richer → spend more. Renters unaffected (or negative: rent inflation). Student cities (Durham, Leeds) — mostly renters, no wealth effect. London/Manchester homeowner venues see +0.2% uplift.
National Living Wage
Annual increase · Apr implementation
+0.4%
NLW £12.21/hr from Apr 2026 (+6.7%). Dual effect: cost push (+£14k/venue/yr labour) vs demand pull (low-wage workers have more to spend). Net impact: slightly positive for revenue, negative for margin. Model separates both signals.
Tax & Duty Changes
Alcohol duty · VAT · business rates
−0.2%
Alcohol duty frozen until Feb 2027. Draught relief extended. VAT stable at 20%. Business rates revaluation: Hull +8%, Manchester −2% (appeals). Duty freeze = no price shock. Any duty increase announcement creates −2% sentiment dip for ~2 weeks.
Events & Entertainment · 14 factors
Football Fixtures
Premier League · Championship · home/away
−1.4%
Newcastle away −4% (fans travel). Man Utd home late KO 20:00 = +8% post-match at Impossible/Factory from 22:30. Hull City home = +12% Tower pre-match (but they go home after). Championship: Leeds home +6% Church. Every fixture tracked with KO time, expected attendance, TV coverage.
DJ & Entertainment Bookings
Headline acts · 12 venues confirmed tonight
+4.8%
Peggy Gou (Digital NCL), Four Tet (Factory MCR). Advance ticket velocity 2.3x median. Tier system: A-list (+40–60%), B-list (+15–25%), resident DJ (baseline). Instagram follower count as demand proxy. Tracked via Resident Advisor, Dice, FIXR, Skiddle APIs.
Competitor Events
Rival venues · festival clashes · arena events
−0.8%
Warehouse Project (MCR) tonight — 18% demo overlap. Mystery shoppers quarterly for pricing. Arena events within 2 miles tracked: AO Arena MCR (Drake, Sat — ends 22:30, spillover to Factory). Utilita Arena NCL shows push crowd to Diamond Strip post-concert.
Concerts & Arena Events
AO Arena · Utilita · First Direct · O2
+2.4%
Drake @ AO Arena MCR tonight (21,000 cap). Ends ~22:30. Post-concert spillover to Deansgate venues: +£8–12k. Newcastle Utilita: Biffy Clyro Fri — indie crowd feeds Digital. Leeds First Direct: nothing tonight (positive — no competition). Tracked via Ticketmaster/AXS APIs.
Music Festivals
Glastonbury · Parklife · Download · Creamfields
−6.2%
Glastonbury (25–29 Jun): drains every UK city −8–12%. 210,000 people in a field. Parklife MCR (13–14 Jun): −22% Factory, −18% Impossible (same demo). Creamfields (Aug bank hol): −15% NCL/MCR. Download (Jun): metal crowd — minimal overlap. Each festival tracked with ticket sales + lineup.
Boxing & PPV Events
Sky/DAZN/TNT · screening venue uplift
+1.8%
No PPV tonight. When active: Tyson Fury fights = +£4–8k per screening venue. Late-night card (23:00+ main event) keeps crowd in venue until 01:00+. Post-fight celebration/commiseration drives extended trading. Also: F1 race weekends (+2–4% at sports bars). UFC London = +6%.
Bank Holidays
Preceding evening · extended weekend
+0%
Next: Mon 5 May. Preceding Sat typically +18–22% (no work Monday). Sunday trading: +35–40% (the “Saturday feeling” on a Sunday). Model pricing in for next Saturday already. Aug bank hol (25 Aug) = biggest: +28% Sat + Creamfields overlap creates regional variation.
Cultural Celebrations
St Patrick’s · Pride · Diwali · CNY · Eid
+0%
St Patrick’s (Mar): Stein Bier Keller MCR +42%, Irish-themed venues +60%. Manchester Pride (Aug bank hol): Canal St + Village +120%, but mainstream venues also +15%. Newcastle Pride (Jul): +18% Diamond Strip. Diwali (Oct): Leicester/Birmingham — not our market currently. Eid al-Fitr (post-Ramadan): Dubai venues +120–150%.
Ramadan
Dubai · 30 days · −40–60% impact
0%
2026: ~27 Feb–28 Mar (ended). Shifts ~11 days/yr (lunar calendar). Dubai venues: −40–60% during, +120–150% Eid week after. No alcohol service during daylight. Late-night iftar creates midnight spike. Model pre-books reduced staff 6 months ahead. Currently: zero impact (Ramadan over).
Sporting Calendar (Non-Football)
Rugby · Cricket · Grand National · Wimbledon
+0.8%
Grand National (12 Apr — just passed): +£3.2k screening venues. Six Nations: +8% across sports pubs on England match days. Cricket: Ashes summers are huge. Wimbledon: minimal direct nightlife impact but outdoor screening events add £2–4k at food/drink venues. Tour de France: zero.
Royal & National Events
Coronations · weddings · funerals · elections
0%
No upcoming events. Royal events create national mood shift: weddings +14% (celebration), funerals −22% (mourning). Elections: −4% on polling day (people stay in watching results). Coronation 2023: +32% pub trade. These are rare but massive when they occur — model has historical precedents.
Halloween
Oct 31 · 2nd biggest night of the year
+38%
2nd only to NYE. 2025 Halloween: Digital NCL +£22k, Impossible MCR +£18k. Advance ticket sales start 6 weeks ahead. Costume theme drives Instagram content (+400% engagement). If Halloween falls on Sat: +45%. If midweek: +22% nearest Saturday. 2026: Saturday 31 Oct = MAXIMUM.
NYE & Christmas Season
Nov–Dec · corporate + consumer peaks
+62%
Christmas party season (late Nov–mid Dec): corporate bookings drive +40–60% at restaurant-bars. NYE: highest rev night of year — Impossible MCR £84k (vs £38k avg Sat). Ticket pricing 3–5x normal. Pre-booking rate: 92%. Model begins NYE marketing allocation in October.
Dry January
Jan health trends · −25% sector-wide
−18%
Worst month of the year. 35% of 18–34s attempt Dry Jan (up from 22% in 2020). First 2 Saturdays worst (−28%), tapers to −12% by late Jan. Model cuts all marketing to zero for w/c 1 Jan, pivots to non-alcoholic promotions. Re-engagement campaign launches 25 Jan. Structural trend: worsening annually.
Transport & Infrastructure · 6 factors
Night Tube (London)
Victoria · Central · Jubilee · Northern · Piccadilly
+0.4%
Running. Night Tube Fri/Sat = +12% late-night trade at Emerald Theatre, One Ninety Four, The Social. When suspended (strikes): −18% after midnight as punters leave early for last regular tube. 00:30 = cliff edge without Night Tube.
Tyne & Wear Metro
Last service times · disruptions
+0.2%
Normal service. Last Metro from Monument: 23:18 (Sat). No night service — forces Uber/taxi after. Strike days: −8% Digital, −12% Neptune (suburban catchment relies on Metro). Engineering works (weekend closures): −4% if Gateshead/South Shields affected.
Rail Strikes & Disruption
ASLEF · RMT · TransPennine · LNER
0%
No strikes scheduled. When active: inter-city venues lose travelling customers. York Impossible −14% on strike days (relies on Leeds/Harrogate rail). TransPennine Express disruption affects MCR↔LDS flow. LNER strikes: −8% Newcastle (London visitors). Model checks RMT/ASLEF calendars daily.
Uber Surge Pricing
Real-time multiplier · late-night barrier
0%
Current surge: 1.0x (normal). On major event nights (NYE, bank holidays): 3–5x surge = punters stay longer (can’t afford to leave) or leave earlier (before surge kicks in). 2.5x+ surge at 01:00 correlates with −8% 01:00–03:00 trade. Tracked via Uber API partner data.
Road Closures & Events
Great North Run · marathon routes · protests
0%
No closures tonight. Great North Run (13 Sep): Newcastle city centre closed — daytime trade +80%, but access issues −4% evening. Manchester Marathon (Apr): Deansgate closed 06:00–16:00. Christmas markets: road closures Nov–Dec but +25% footfall overall. Model checks council event calendars.
Parking & Access
NCP availability · event-day restrictions
0%
Normal. NCP Deansgate: 420 spaces. St James’ Park match days: 2,000 spaces restricted — Digital/Tup Tup lose drive-in trade. Ibiza: no parking relevance (taxi/walk culture). Dubai: valet standard — parking never a barrier. Only factors into venues with driving demographics.
Social & Digital Signals · 8 factors
Instagram Engagement
Per venue · story views · reel reach
+1.4%
Digital NCL reel: 84k views (4.2x avg). Peggy Gou announcement post = 12.4k saves. Saves-to-attendance correlation: 0.78. When announcement post gets >5k saves, model increases baseline by +8%. Instagram story poll (“Coming tonight?”) responses = real-time demand signal.
TikTok Virality
Venue tags · trend hooks · UGC volume
+2.2%
Impossible MCR cocktail video: 2.1M views this week (UGC). #ImpossibleManchester tag volume +340% WoW. TikTok viral moments create 2–4 week demand waves. Last viral (ice bucket challenge Jan): +£14k over 3 weekends. Model detects >500k view spikes and adjusts baseline +4–6%.
Google “Near Me” Searches
“Bars near me” · “clubs tonight” · by city
+0.6%
“Bars near me Manchester” +18% vs 4-week avg (Saturday). “Clubs tonight Newcastle” +12%. Real-time intent signal. Correlates with walk-in traffic 2–4 hours later. Searchconsole data + Google Trends API. Also tracking: “queue times Digital Newcastle”, “dress code Tup Tup”.
Ticket Pre-Sales
FIXR · Skiddle · Fatsoma · Dice · RA
+3.8%
Tonight across group: 4,820 advance tickets sold (vs 3,200 avg Saturday). Digital NCL: 1,240 (Peggy Gou — sellout). Factory MCR: 680 (Four Tet — 78% sold). Ticket velocity (sales per hour) is best real-time predictor. Velocity >2x median = model adds +12% to forecast.
Google Reviews & Ratings
Score changes · review volume · sentiment
0%
Impossible MCR: 4.3☆ (stable). Tower Hull: 3.1☆ (−0.4 in 90 days — concerning). Every −0.5☆ drop correlates with −8% walk-in traffic over following 60 days. Review keywords tracked: “rude staff”, “overpriced”, “long wait” trigger operational alerts. Volume spike = recent event (good or bad).
Influencer & Press Mentions
Print · online · radio · podcast
+0.4%
Manchester Evening News feature on Impossible MCR terrace (Tue). Estimated reach: 180k. Emerald Theatre review in TimeOut London: 8.2k reads. DJ Mag feature on Digital NCL: 22k. Tracked via Meltwater/Brandwatch. Press mentions have 7–14 day decay curve on footfall.
Competitor Social Activity
Rival venue announcements · pricing · promos
−0.2%
Warehouse Project announced Skrillex (same night as Four Tet @ Factory). 14k engagement on post. Estimated cannibalisation: 240 covers. But different demos: WHP = bassline/dubstep, Factory = electronic/house. Cross-demo overlap ~12%. Model uses genre tagging to weight cannibalisation.
WhatsApp & Group Chat Signals
Promoter list activity · booking enquiry volume
+1.2%
Promoter WhatsApp lists: 14,200 subscribers across group. Tonight’s blast sent 16:00 — open rate 68% (vs 42% avg). Table booking enquiries via WhatsApp: 34 (vs 18 avg Saturday). Enquiry-to-booking conversion: 72%. High WhatsApp activity = model adds +2–4% to walk-in forecast.
Calendar & Seasonal · 6 factors
Day of Week & Position in Month
1st Sat vs last Sat · mid-month trough
base
Last Saturday of April = payday alignment. 1st Saturday = +4% (post-payday glow). 2nd Saturday = baseline. 3rd Saturday = −2% (mid-month squeeze). Last Saturday = +6% (payday). Consistent pattern across 3 years of EPOS data. Friday = 62% of Saturday revenue on average.
Clocks Change (BST/GMT)
Spring forward · Autumn back · trading hour shift
+0%
Spring forward (30 Mar): lose 1 trading hour — −£2–4k per venue. Autumn back (26 Oct): gain 1 hour — +£3–5k. Spring forward Saturday night is the only Saturday with 23 hours — model adjusts baseline down. Ibiza: +1hr CET vs UK, but irrelevant (venues trade until 07:00).
Summer Solstice
21 Jun · longest day · outdoor peak
+4.2%
Sunset 21:38 (Manchester). Extended terrace hours. Ibiza: sunset 21:28 — golden hour revenue peak. Outdoor venue revenue peaks within ±2 weeks of solstice. Longest daylight = highest footfall probability. If solstice Saturday is dry: historic best outdoor trading day of the year.
Winter Solstice / Dark Evenings
Dec · sunset 15:52 · seasonal affect
−3.8%
Dec Saturdays: sunset 15:52. Dark by 16:30. SAD (Seasonal Affective Disorder) prevalence: 6% of UK adults. Going-out reluctance peaks mid-Nov to mid-Jan. Offset by Christmas party season (Nov–Dec) but underlying demand is lower. Jan+Feb = structural trough.
Valentine’s Day
14 Feb · restaurant premium · club drag
split
Restaurant-bars (Impossible, One Ninety Four): +28% (couples dining, prix fixe menus). Nightclubs: −8% (couples stay in). If 14 Feb falls on Saturday: maximum restaurant uplift (+42%), maximum club drag (−14%). 2026: Saturday 14 Feb — model pre-allocated restaurant marketing, cut club spend.
Coachella / Ibiza Opening Season
Global nightlife calendar · aspirational demand
+1.8%
Ibiza opening parties (May): global attention on nightlife. Instagram content volume peaks. UK venues benefit from aspirational demand — people who can’t go to Ibiza go out locally instead. Coachella (Apr, USA): Desert Hills PSP +44%. Cultural tentpole events create sector-wide uplift.
Business & Corporate Events · 10 factors
Trade Exhibitions & Expos
NEC · Manchester Central · ExCeL · Harrogate
+6.4%
Major exhibitions bring 5,000–80,000 delegates to host cities for 2–5 days. Hotel occupancy spikes, after-work spending peaks Wed/Thu. Manchester Central events: +£12k at Impossible/BLVD on expo evenings. NEC events feed Birmingham (no venues yet) but delegates travel to MCR. Tracked via Eventbrite + exhibition calendars.
Legal & Solicitors Conferences
Law Society · SRA · regional bar events
+4.8%
Legal conferences in Manchester (Midland Hotel, Hilton Deansgate) = high-spend corporate groups. Avg spend £68/head vs £42 regular. Law Society annual conference (Oct): +£18k across MCR venues. Regional bar dinners: quarterly, 200–400 covers. Thursday nights are peak — after-dinner migration to BLVD, Impossible VIP.
Government & Political Events
Party conferences · council events · civil service
+3.2%
Conservative Party Conference (Manchester, Oct): +£28k group revenue over 4 days. 12,000 delegates, journalists, lobbyists. Expense-account spending peaks. Labour Conference (Liverpool, Sep): +£14k at Magnet + local venues. Council away days and civil service conferences: smaller but consistent Wed/Thu uplift in host cities.
Motor Shows & Car Launches
Dealer events · press launches · auto expos
+2.8%
Car brand dealer conferences: BMW (annual, 800 delegates), Mercedes launch events, JLR press days. High-spend male demographic 30–50. After-event spending at premium venues: avg £85/head. Brands often book private dining at Impossible/One Ninety Four. Tracked via automotive industry calendar + direct venue enquiries.
Financial & Banking Events
HSBC · Barclays · Deloitte · PwC · KPMG
+5.2%
Big 4 accountancy firms and banks hold quarterly socials, summer parties, Christmas events. Deloitte MCR (2,500 staff) = £40k annual private bookings at Impossible alone. HSBC team nights: monthly, 80–120 covers. Dubai DIFC = financial district — Thursday night is “brunch” (corporate letting loose). Avg spend 3.2x regular.
Medical & Pharma Conferences
BMA · NHS Trust events · pharma launches
+2.4%
Manchester is the UK’s second-largest medical conference city. BMA annual (Jun, 2,000 delegates). Pharma product launches at Hilton/Midland. NHS payday (27th) + conference = double uplift. Doctors/consultants avg spend £62/head. Newcastle: RVI and Freeman Hospital generate consistent local medical trade at Digital/Tup Tup.
Tech & Startup Conferences
Web Summit · TechNorth · DWP Digital · AWS
+1.8%
Manchester Digital (2,400 members) holds monthly events. AWS Summit (Manchester Central, May): 5,000 attendees, heavy after-party culture. Leeds Digital Festival (Sep): +£8k Church/Get Baked. Tech crowd skews younger (25–35), prefers cocktail bars. Sponsors often book private hire — tracked via Eventbrite + LinkedIn Events.
Wedding Season & Private Hire
May–Sep peak · venue buyouts · after-parties
+3.6%
Impossible MCR: 42 wedding bookings in 2026 (May–Sep). Avg spend £18k per event. Saturday buyouts = venue closed to public but guaranteed £15–25k. After-party spillover to BLVD/Factory. York Impossible: Grade II listed = premium wedding venue, 28 bookings. Model blocks venue from forecast on buyout days and redirects walk-in marketing to nearby venues.
Gentlemen’s Club Drivers
Stag dos · corporate entertaining · race days
+8.4%
Whiskey Down venues correlate with: horse racing (Cheltenham +62%, Aintree +48%, Royal Ascot +34%), stag party weekends (tracked via DesignMyNight group bookings), corporate entertainment nights (law firms, finance, car dealers). Exhibition weeks: +22% at gentlemen’s venues (delegates on expenses). Key dates: Cheltenham Gold Cup (Mar), Grand National (Apr), York Races (Aug). Thursday/Friday peaks. Avg spend £120/head vs £42 standard nightlife.
Awards Ceremonies & Galas
Industry awards · charity galas · black-tie events
+2.2%
Manchester hosts 40+ annual industry awards (Business Masters, E-Commerce Awards, Northern Power Women). Each puts 300–800 black-tie guests into the city centre from 22:00. After-party culture drives premium venue trade: champagne spend +240%, VIP table bookings +180%. Newcastle: North East Business Awards (Apr, 600 guests). These events are booked 6–12 months ahead — model ingests them early.
07·c
Confirmed ROAS attribution
Baseline vs Marketing Uplift · TonightWhere the marketing £ went
| Venue | Baseline | Actual | Uplift | Spend | ROAS | Verdict |
|---|---|---|---|---|---|---|
The Emerald Theatre LONDON W1 | £18,200 | £28,640 | +£10,440 | £4,800 | 2.18x | SCALE |
Digital NEWCASTLE | £26,100 | £32,180 | +£6,080 | £3,200 | 1.90x | SCALE |
Impossible MANCHESTER | £34,800 | £38,420 | +£3,620 | £2,400 | 1.51x | HOLD |
Hangar 8289 IBIZA | £14,200 | £16,920 | +£2,720 | £1,800 | 1.51x | HOLD |
Tup Tup Palace NEWCASTLE | £23,200 | £24,750 | +£1,550 | £800 | 1.94x | SCALE |
Church LEEDS | £11,400 | £9,840 | −£1,560 | £600 | −2.60x | REVIEW |
Tower Ballroom HULL | £7,200 | £6,920 | −£280 | £200 | −1.40x | RELAUNCH |
07·d
Autonomous budget allocation
AI Revenue Optimising AgentsNext Saturday · recommended reallocation
Scale: Digital Newcastle — +£1,800 Meta Ads
ROAS 1.90x this week. Peggy Gou content performing 3.2x CTR benchmark. Agent recommends lookalike audience expansion on Instagram Reels. Predicted uplift: +£4,200.
AUTO
Scale: Tup Tup Palace — +£1,200 FixR + Instagram
High ROAS (1.94x) but underspent at £800/wk. Predicted 25 Apr baseline strong (payday + dry weather). Agent pushing early-bird ticket campaign. Predicted uplift: +£2,800.
AUTO
Relaunch: Tower Ballroom Hull — £2,500 rescue campaign
Zero organic pull. Agent creating TikTok + Instagram campaign: £5 entry + free drink before midnight. Targeting Hull 18–30 demographic. Weather dry — outdoor queue content. Predicted recovery: £12k (vs £6.9k current).
PENDING
Reduce: Impossible MCR — −£800 (diminishing returns)
At £2,400/wk spend, hitting ceiling. Venue already at record £38k. Marginal ROAS dropping to 1.2x. Redirect to underperforming venues with higher headroom. Agent reallocating to Church Leeds + Tower Hull.
AUTO
Pause: Church Leeds — fix ops before spending
Negative ROAS −2.60x. Marketing isn’t the problem — labour at 31.4%, voids at 284/60d, cash variance £180. Agent recommends: fix ops first, then relaunch with £1,500 student night campaign once voids resolve.
HOLD
Weekly budget · auto-allocated
£14.2k
across 68 venues · 4 channels
Predicted group ROAS
2.4x
vs. 1.8x manual allocation LM
Predicted uplift vs baseline
£60k
next Saturday · from £468k → £528k
07·e
How margin increases
This is not a dashboardSix ways Pulse increases EBITDA
1
Labour — cuts cost in real-time
Church Leeds tonight: labour at 31.4% vs 24% target. That’s £2,900 wasted. Pulse detected it at 23:28, notified the GM to stand down 2 staff. That’s £180 saved in one night, one venue. A dashboard shows you the problem on Monday. Pulse fixes it on Saturday night.
£340k
ANNUAL SAVING
2
Shrinkage — stops stock walking out the door
Tower Ballroom: 380 voids in 60 days, £300 cash variance, cross-correlated. That’s not a reporting insight — that’s Pulse flagging probable theft and triggering a GM investigation before the next shift. Group shrinkage dropped from 4.2% to 2.8%.
£512k
ANNUAL SAVING
3
Marketing — same spend, more revenue
Impossible MCR got £2,400 and hit a ceiling (1.51x ROAS). Tower Hull got £200 and died. The engine takes £800 from Impossible (diminishing returns) and puts it into Tower Hull (headroom). Same total spend. More total revenue. More revenue on the same rent and rates = higher margin.
£2.2M
INCREMENTAL GP
4
Predictive staffing — rotas built on data, not gut feel
Model says next Tuesday will be quiet: mid-month, rain forecast, no football. So you rota 8 fewer staff across the group. A report tells you last Tuesday was overstaffed. Pulse makes sure next Tuesday isn’t.
£220k
ANNUAL SAVING
5
Cash handling — catches variance before it compounds
£300 cash variance at one venue for one night is a flag. £300/night over 30 nights is £9,000. Pulse cross-references voids + cash + timestamps and escalates on night one, not day thirty.
£180k
LOSS PREVENTED
6
Revenue on fixed costs — the multiplier
The marketing engine drives £60k additional revenue per week. Rent doesn’t change. Rates don’t change. Insurance doesn’t change. That £60k drops through at 71.2% GP = £42,700 extra gross profit per week on zero additional fixed overhead.
£2.2M
ANNUAL GP UPLIFT
Total annualised EBITDA impact
£3.4M
£340k labour + £512k shrinkage + £220k staffing + £180k cash + £2.2M GP uplift
Current group EBITDA · run rate
£10.4M
Pulse impact = +32.7% EBITDA uplift on same venue estate
08
Revenue heatmap
Top Venue · Sat
£38.4k
Impossible MCRAll-time Saturday record
Group Avg · per venue
£7.2k
↑ +8.4% vs. LWMedian £9.8k
Best Day · This Week
Saturday
£487k group totalFri £412k · Thu £284k
Weakest Venue · 7d
£3.8k
Get Baked LDSvs. £14.8k BE target
Daily Revenue by VenuePerformance intensity
♦ Trading Hours Analysis · Burn Rate Engine
When to open & close
Every hour a venue is open costs money: staff, security, utilities, DJ fees. This engine analyses minute-by-minute EPOS data against labour cost to find the exact hours where you’re making money and the hours where you’re burning it. The result: optimised opening and closing times that maximise profit per trading hour.
15·a
Hourly burn rate · Impossible MCR
Revenue vs Cost per HourWhen is the venue making vs burning money?
Recommended: Open 20:00 · Close 03:30
Currently opens 18:00 (2 hours early burn: −£600). Stays open until 05:00 (1.5 hours late burn: −£700). Shifting to optimal window saves £1,300 per Saturday = £67,600/year at this venue alone.
£67.6k
15·b
Hourly P&L breakdown
| Hour | Revenue | Staff Cost | Door Cost | DJ/Sound | Utilities | Total Cost | Net P&L | Covers | £/Head | Status |
|---|---|---|---|---|---|---|---|---|---|---|
| 18:00 | £240 | £286 | £202 | £80 | £52 | £620 | −£380 | 12 | £20 | BURN |
| 19:00 | £480 | £342 | £202 | £80 | £76 | £700 | −£220 | 28 | £17 | BURN |
| 20:00 | £880 | £342 | £269 | £120 | £69 | £800 | +£80 | 48 | £18 | B/E |
| 21:00 | £2,220 | £342 | £269 | £120 | £69 | £800 | +£1,420 | 98 | £23 | PROFIT |
| 22:00 | £5,280 | £456 | £336 | £120 | £88 | £1,000 | +£4,280 | 186 | £28 | PROFIT |
| 23:00 | £7,840 | £456 | £336 | £120 | £88 | £1,000 | +£6,840 | 248 | £32 | PEAK |
| 00:00 | £9,200 | £456 | £336 | £120 | £88 | £1,000 | +£8,200 | 312 | £29 | MAX |
| 01:00 | £6,440 | £342 | £269 | £120 | £69 | £800 | +£5,640 | 224 | £29 | PROFIT |
| 02:00 | £3,400 | £228 | £202 | £120 | £50 | £600 | +£2,800 | 142 | £24 | PROFIT |
| 03:00 | £920 | £228 | £202 | £120 | £50 | £600 | +£320 | 48 | £19 | TAPER |
| 04:00 | £320 | £228 | £202 | £120 | £50 | £600 | −£280 | 18 | £18 | BURN |
| 05:00 | £180 | £228 | £202 | £120 | £50 | £600 | −£420 | 8 | £23 | BURN |
| TOTAL | £37,400 | £3,934 | £3,027 | £1,360 | £849 | £9,170 | +£28,230 | 1,372 | £27 | 75.5% GP |
15·c
Optimal hours across the group
If every venue adopted optimal hoursGroup annual saving: £842,000
| Venue | Current Open | Current Close | Optimal Open | Optimal Close | Early Burn | Late Burn | Sat Saving | Annual |
|---|---|---|---|---|---|---|---|---|
Impossible MANCHESTER | 18:00 | 05:00 | 20:00 | 03:30 | −£600 | −£700 | £1,300 | £67,600 |
Digital NEWCASTLE | 22:00 | 05:00 | 22:00 | 04:00 | £0 | −£480 | £480 | £24,960 |
Factory (FAC251) MANCHESTER | 22:00 | 06:00 | 22:30 | 04:30 | −£180 | −£620 | £800 | £41,600 |
Tup Tup Palace NEWCASTLE | 22:00 | 04:30 | 22:00 | 03:30 | £0 | −£540 | £540 | £28,080 |
Church LEEDS | 22:00 | 05:00 | 22:30 | 03:00 | −£220 | −£880 | £1,100 | £57,200 |
Klute DURHAM | 21:00 | 03:00 | 22:00 | 02:30 | −£340 | −£180 | £520 | £27,040 |
Tower Ballroom HULL | 21:00 | 04:00 | 22:30 | 02:00 | −£480 | −£940 | £1,420 | £73,840 |
| 7 venues sampled | −£1,820 | −£4,340 | £6,160 | £320,320 | ||||
| Projected: all 49 UK venues | Extrapolated using per-venue avg · conservative estimate | £16,200/wk | £842,400 | |||||
15·d
Trading pattern vs events
How Events Shift the Trading WindowThe same venue trades differently every Saturday
DJ headline night (Peggy Gou @ Digital): Peak shifts +1hr later
Regular Saturday peak: 23:00–00:00. DJ night peak: 00:00–01:30. Crowd arrives later, stays later. Optimal close shifts from 04:00 to 04:30. Open time stays 22:00. Net: +£2.4k revenue but +30 min staff cost. ROI: 4.2x.
+1HR
Football late KO (Man Utd 20:00): Post-match surge at 22:30
Normal 22:00 hour: £1,200. Post-match 22:00 hour: £3,800 (+217%). Crowd arrives in a wave, not gradually. Staff should be pre-positioned by 22:15. Kitchen close can extend 30 min (food demand spikes post-match).
+217%
Rain forecast: Crowd arrives 45 min earlier, leaves 30 min earlier
Dry Saturdays: peak footfall 23:00. Rainy Saturdays: peak footfall 22:15 (people don’t want to queue in rain, arrive early). But also leave by 02:30 vs 03:30. Net: same revenue, compressed window. Close 30 min earlier on rain nights.
−30M
Payday Saturday: Avg spend +£6/head, peak extends +45 min
Regular close-profitable: 03:00. Payday close-profitable: 03:45. People spend more per round AND stay later. Optimal: keep full staff 45 min longer on payday weekends. Extra 45 min revenue: £1,800 at Impossible (well above cost).
+45M
Exam season: Peak drops 40%, optimal close pulls back 1.5 hours
Durham exam weeks: Klute peak drops from £4.2k/hr to £2.5k/hr. Break-even hour shifts from 02:30 to 01:00. Recommended: close at 01:30 during exams (currently stays open until 03:00 = 1.5hrs burn). Saves £780/night.
−1.5HR
15·e
Drinks intelligence by hour
What’s selling when — and why it mattersCategory mix shifts through the night
| Hour | Draught | Spirits | Cocktails | Wine | Champagne | Soft/NA | Food | Avg £/Txn | Category Mix |
|---|---|---|---|---|---|---|---|---|---|
| 18:00 | 42% | 8% | 12% | 22% | 0% | 4% | 12% | £8.40 | |
| 21:00 | 34% | 18% | 22% | 12% | 4% | 6% | 4% | £12.80 | |
| 23:00 | 18% | 28% | 24% | 6% | 14% | 8% | 2% | £18.60 | |
| 00:00 | 12% | 32% | 18% | 4% | 22% | 10% | 2% | £24.40 | |
| 02:00 | 8% | 38% | 14% | 2% | 18% | 16% | 4% | £16.20 |
Draught
Spirits
Cocktails
Wine
Champagne
Soft/NA
Food
AI InsightWhen does the champagne start flowing?
Champagne sales spike at 23:15 — triggered by VIP table occupation
First champagne order correlates 0.89 with VIP table being seated. Average time from table seating to first bottle: 18 minutes. If VIP tables fill before 23:00, champagne hour starts 22:45. Model tracks VIP check-in time → predicts champagne window → pre-chills stock.
23:15
Promo sparkler shows increase table spend by 34% within 20 minutes
When bar staff do sparkler/bottle presentations at VIP tables, surrounding tables order bottles within 20 minutes 34% more often (FOMO effect). Optimal sparkler timing: first at 23:30 (sets the tone), then every 45 min. Cost per sparkler show: £2.40. Average triggered spend: £280. ROI: 117x.
117x ROI
Spirit-to-cocktail ratio predicts crowd mood and close time
When spirits >35% of sales (shots culture), crowd is high-energy and stays later (+30 min avg close). When cocktails >25%, crowd is more measured — higher avg spend but leaves on time. Model uses drink ratio at 23:00 as a leading indicator for 02:00+ trading viability. Shots crowd = keep full staff; cocktail crowd = begin staff drawdown at 01:30.
PREDICTOR
Recommended: Start first promo show at 23:30, not 00:30
Currently, Impossible MCR does first VIP sparkler show around 00:30. Data shows 23:30 first show triggers 42% more follow-on purchases (crowd still arriving, social proof peaks). Moving first show 1 hour earlier = estimated +£1,800 champagne revenue per Saturday = £93,600/year.
£93.6k
15·f
Weekly problem tracker
Week of 14–19 April 2026Problems identified · solutions recommended
Tower Hull: 380 voids in 60 days + £300 cash variance
AI Recommendation: Cross-reference void timestamps with CCTV. Pattern suggests systematic issue at Terminal 3 (Thu/Sat 22:00–01:00). Recommend: GM investigation + till reassignment. Owner: Dave Mitchell (GM) · Due: 21 Apr
URGENT
Church Leeds: Labour at 31.4% (target 24%)
AI Recommendation: Over-rostered by 3 staff on Saturdays. Revenue doesn’t support current team size. Reduce Sat rota from 14 to 11 bar staff. Saves £540/wk. Also: open 30 min later (22:30 vs 22:00). Owner: Sarah Collins (AGM) · Due: 22 Apr
HIGH
Klute Durham: Revenue −8.2% WoW during term time
AI Recommendation: Reading week starts 28 Apr — expect further −22%. Pre-action: cut marketing to zero for w/c 28 Apr, reduce staff by 4. Post reading week: launch “welcome back” campaign with £3 drinks promotion. Owner: Dillon Hayes (GM) · Due: 25 Apr
HIGH
Impossible MCR: Early burn £600/Sat (opening at 18:00)
RESOLVED 16 Apr: Aaron approved shift to 19:30 open (GM + 2 bar staff only until 20:00, door team from 20:00). First Saturday saving: £480. Tracking for 4 weeks then permanent change.
DONE
Digital NCL: Champagne promo starting too late (00:30)
RESOLVED 17 Apr: First sparkler show moved to 23:30. Week 1 result: champagne revenue +28% (£4,200 → £5,376). Confirmed as permanent change.
DONE
P
Pulse AI
AI Command Centre · Always online
09
EPOS live tills & cash-up
Live tills · active session
Transaction feed
LIVE
Cash-up summary
Operator leaderboard
| # | Operator | Role | Txns | Revenue | Avg Txn | Voids | Speed |
|---|
Dojo online payments
♦ Inbound Intelligence · Global Flow Engine
Tourist flow predicts revenue
ML model tracks passenger flows between 30 origin cities and 14 global destinations. When 4,000 18–30 year olds fly out of Manchester to Ibiza, Manchester clubs lose £136k of potential revenue. Ibiza gains it at 2.5x spend. The following weekend, Manchester bounces +5–8% from FOMO. We measure every link in that chain.
11·a
Three-layer intelligence architecture
ML
Prediction Engine
Proprietary ML ensemble trained on historical revenue × 70+ factors. Predicts baseline per venue for 15 days. Runs attribution (actual − baseline = marketing uplift). Retrains weekly. 68 venues × 15 days in <2 seconds.
ACCURACY: 96.9% · RETRAIN: WEEKLY
LLM
Reasoning Layer
Multi-provider LLM layer interprets ML outputs, explains WHY predictions are what they are. Generates natural language briefings. Answers CEO questions in Command chat. Writes marketing copy for autonomous deployment.
PROVIDERS: 6 · FAILOVER: AUTO
AGT
Autonomous Execution
AI agents take ML predictions + LLM reasoning and execute: reallocate ad budgets, create campaigns, notify GMs, adjust staffing, trigger content creation. Closed loop — results feed back into ML training.
AGENTS: 25+ · LOOP: CLOSED
11·b
Live flight flows · this weekend
Bidirectional Flow MatrixWhere your customers are flying
| Route | Flights | Pax (est.) | 18–34 % | Avg Fare | Seat Load | Revenue Drain | Dest. Gain | Impact |
|---|---|---|---|---|---|---|---|---|
| Outbound · Manchester Airport (MAN) | ||||||||
Manchester → Ibiza Ryanair FR1842, easyJet U28421, Jet2 LS801, +3 | 6 | 1,240 | 72% | £89 | 94% | −£42.2k | +€76.0k | MCR ↓ |
Manchester → Mallorca Jet2 LS303, TUI BY4842, easyJet U28601 | 3 | 580 | 54% | £112 | 88% | −£14.8k | +€28.4k | MCR ↓ |
Manchester → Hvar (via Split) easyJet U28911 | 1 | 180 | 58% | £134 | 82% | −£6.1k | +€15.3k | MCR ↓ |
Manchester → Dubai Emirates EK18, TUI BY426 | 2 | 640 | 32% | £420 | 91% | −£8.2k | +AED 82k | MCR ↓ |
Manchester → Alicante Ryanair FR6214, Jet2 LS921, easyJet U28112 | 3 | 620 | 42% | £78 | 96% | −£9.8k | — | older demo |
Manchester → Faro Ryanair FR8224, Jet2 LS641 | 2 | 380 | 38% | £92 | 90% | −£5.4k | — | older demo |
Manchester → Tenerife TUI BY2648, Jet2 LS1121 | 2 | 420 | 36% | £148 | 92% | −£5.6k | — | minimal |
| Outbound · Newcastle Airport (NCL) | ||||||||
Newcastle → Ibiza Jet2 LS1801, Ryanair FR4412, Ryanair FR4414 | 3 | 620 | 68% | £98 | 96% | −£21.1k | +€38.0k | NCL ↓ |
Newcastle → Alicante Jet2 LS911, Ryanair FR6208 | 2 | 440 | 44% | £82 | 94% | −£7.2k | — | older demo |
Newcastle → Mallorca Jet2 LS313 | 1 | 180 | 62% | £118 | 86% | −£5.8k | +€12.4k | NCL ↓ |
Newcastle → Faro Jet2 LS651 | 1 | 189 | 34% | £96 | 88% | −£2.4k | — | minimal |
Newcastle → Dubai Emirates EK152 (seasonal, ends May) | 1 | 340 | 28% | £480 | 78% | −£3.8k | +AED 44k | premium |
| Outbound · Leeds Bradford (LBA) | ||||||||
Leeds → Mallorca Jet2 LS305, TUI BY6482 | 2 | 380 | 64% | £105 | 90% | −£12.9k | +€23.2k | LDS ↓ |
Leeds → Ibiza Jet2 LS1803 | 1 | 189 | 74% | £108 | 98% | −£6.8k | +€14.2k | LDS ↓ |
Leeds → Faro Ryanair FR8216 | 1 | 189 | 42% | £74 | 84% | −£3.2k | — | minimal |
| Outbound · London (LHR/LGW/STN/LTN) | ||||||||
London → Dubai Emirates ×7, BA ×4, Virgin ×2, Qatar ×1 | 14 | 3,800 | 38% | £380 | 89% | −£49.1k | +AED 184k | LDN ↓ |
London → Ibiza easyJet ×8, Ryanair ×6, BA ×4, Vueling ×4 | 22 | 4,180 | 66% | £72 | 92% | −£18.4k | +€142k | diluted |
London → LA (LAX) BA ×3, Virgin ×2, AA ×2, Norse ×1 | 8 | 2,400 | 34% | £520 | 86% | −£12.2k | +$18.6k | diluted |
London → NYC BA ×6, Virgin ×3, JetBlue ×2, Delta ×2 | 13 | 3,600 | 42% | £440 | 84% | −£14.8k | +$22.4k | diluted |
| Inbound · Destination Arrivals | ||||||||
UK → Ibiza (all airports) IBZ arrivals from MAN, NCL, LGW, STN, LTN, BRS, BHX | 42 | 8,240 | 64% | £92 | 93% | — | +€280k | IBZ ↑ |
Germany → Ibiza FRA, MUC, DUS, HAM, BER · Condor, Eurowings, Lufthansa | 12 | 2,800 | 58% | €108 | 88% | — | +€94k | IBZ ↑ |
UK → Dubai (all airports) Emirates, BA, Virgin · LHR, MAN, NCL, BHX | 18 | 5,200 | 34% | £410 | 87% | — | +AED 310k | DXB ↑ |
Split → Hvar (ferry) Jadrolinija · Krilo catamaran · 4 daily | 4 | 480 | 52% | €22 | 74% | — | +€32k | HVR ↑ |
| FOMO Bounce · Following Weekend Return Effect | ||||||||
FOMO Bounce → Manchester Post-Ibiza Instagram effect · following weekend | — | +820 | 74% | — | — | +£27.9k | — | MCR ↑ |
FOMO Bounce → Newcastle Post-Ibiza · Instagram stories from NCL demo | — | +440 | 72% | — | — | +£14.8k | — | NCL ↑ |
FOMO Bounce → Leeds Post-Mallorca · student demo return | — | +280 | 68% | — | — | +£9.4k | — | LDS ↑ |
Total outbound pax today
24.8k
from 6 UK airports · 89 flights
18–34 drain (our demo)
12.4k
50.2% of total pax in target demo
Net UK revenue drain
£242k
offset by FOMO bounce: +£52k next wknd
Destination venue gain
€682k
Ibiza + Mallorca + Dubai + Hvar
11·c
Destination intelligence
Ibiza
Season: May–Oct · 3 Tokyo venues
€37.6k
42,000 arrivals this month (+10.5% YoY). 64% aged 18–34. 41% British, 22% German. Payday weekend + Cirque residency at AMOK. Opening party season in full swing.
Mallorca
Season: May–Oct · 1 Tokyo venue (AMOK)
€10.8k
38,000 arrivals. 48% aged 18–34 (older skew than Ibiza). 35% German, 28% British. Magaluf = 18–25, Palma = 25–40 split. AMOK is Palma demographic.
Dubai
Season: Oct–Apr · 2 Tokyo venues
AED 62.3k
Peak season. 28°C. Post-Ramadan spending strong. UK 12%, India 15%, Saudi 10%, Russia 8%. DSF ended Feb — shoulder period before summer dead zone.
Hvar, Croatia
Season: Jun–Sep · 1 Tokyo venue
€6.8k
Early season ramp. Split airport +14% YoY. Ultra Europe (early Jul) drives spillover. Yacht charter bookings running +22% ahead of pace. Premium demographic.
11·d
70+ factors per venue per night
Revenue DecompositionImpossible Manchester · Sat 25 Apr prediction
Seasonal Baseline
April Saturday, good month, post-Easter, pre-summer
£32,400
base
Weather
Dry, 16°C, sunset 20:22. Beer garden capacity active
+£2,100
+6.5%
Payday
Last Friday of month. Avg spend +£4–6 per head
+£3,200
+9.9%
Football
Man Utd home, late KO 20:00. Post-match crowd from 22:30
+£1,800
+5.6%
Outbound Flight Drain
6 MCR→IBZ flights today. 1,240 pax in target demo left the city
−£1,400
−4.3%
FOMO Effect
High Ibiza Instagram volume this week. Going-out intent elevated
+£800
+2.5%
Competitor Events
Warehouse Project at Depot Mayfield. 18% demo overlap
−£600
−1.9%
DJ Booking
Four Tet headline. Advance ticket velocity 2.3x median
+£2,400
+7.4%
Macro (Wage Growth + Confidence)
Real wages +1.2%, GfK +2. Consumer spending recovering
+£400
+1.2%
Marketing (Autonomous Engine)
£2,400 spend via Debut engine. Instagram + FixR. Predicted 1.5x ROAS
+£3,600
+11.1%
Predicted Revenue
Confidence: 97.2% · Tolerance: ±2.4%
£44,700
+37.9%
11·e
Seasonal calendars by region
UK Nightlife
49 venues · year-round with peaks
Peak: NYE, Halloween, Freshers’ wk, Bank holiday eves, Xmas parties (Nov–Dec)
Watch: Dry January, exam seasons, festival weekends (Glastonbury drains cities)
Low: Early Jan, early Feb, mid-week Aug (families away)
Events: St Patrick’s (+40%), Pride weekends, graduation weeks, Easter Sat
Ibiza / Mallorca
4 venues · May–October
Peak: Jul–Aug (100%). Opening parties (May). Closing parties (Sep–Oct)
Shoulder: Jun (70–85%), Sep (75–90% — best weather, wealthier demo)
Dead: Nov–Mar (0–10%). Most venues close entirely
Key dates: Opening/closing party nights = 2–3x normal pricing
Dubai
2 venues · inverted + religious calendar
Peak: Dec–Jan (NYE biggest night). Oct–Nov (season opens, F1 Abu Dhabi)
Dead: Jun–Aug (45–50°C). Ramadan (−40–60%)
Spikes: Eid al-Fitr (+120–150%), Eid al-Adha, Dubai Shopping Festival
Note: Ramadan shifts ~11 days/yr. 2026: ~27 Feb–28 Mar
Croatia (Hvar)
1 venue · Jun–Sep
Peak: Jul–Aug. Ultra Europe festival (Split, early Jul) = spillover
Shoulder: May (yacht preview), Sep (older/wealthier, grape harvest events)
Dead: Nov–Apr. Island pop drops 11k → 3.5k. Ferries reduced
Premium: Yacht crowd. Avg spend €120/hd vs €85 Ibiza
USA (LA / Palm Springs / NYC)
4 venues · mixed seasonal
LA: Year-round. Awards season (Jan–Mar) premium. Coachella (Apr) empties LA
Palm Springs: INVERTED. Peak Oct–May (desert winter). Dead Jun–Sep (45°C+)
NYC: Fashion weeks (Feb, Sep) = premium. Hamptons exodus kills Jul–Aug weekends
Dead: Thanksgiving week (nationwide). Post-July 4th lull
Global Expansion Targets
14 destinations tracked · ready to onboard
Tracked: Mykonos, Barcelona, Amsterdam, Tulum, Bali, Las Vegas, Miami, Cancun
Also: Zante, Ayia Napa, Tenerife, Marbella, Lisbon, Berlin
Ready: Same model architecture. Plug in EPOS + local factors = live in 48h
Moat: First platform to correlate origin↔destination flows globally
11·f
Data sources & training
ML Training Data PipelineHistorical + real-time + forward-looking
Eurostat Aviation API — free, REST, monthly
Passenger counts by route pair (e.g., Manchester→Ibiza). All EU/EEA airports. 10+ years of history. ML training backbone for seasonal + flow models.
LIVE
Ibestat Tourist Profiles — free, quarterly, demographics
Age (16–24, 25–34, 35–44), sex, nationality, spend/day, activities. Ibiza + Mallorca. The gold source for nightclub demographic profiling.
LIVE
Google Trends API — free, weekly, forward-looking
“Flights to Ibiza” search volume from Manchester IPs. 4–8 week leading indicator for actual passenger numbers. Proxy for ForwardKeys at zero cost.
LIVE
EPOS Historical Data — from Tokyo Industries
Zonal EPOS exports. Daily revenue per venue going back 3+ years. The core training dataset. Combined with external factors = 70+ feature ML model.
PENDING
OAG Scheduled Capacity — paid, daily, 12 months ahead
Every flight in every airline’s schedule. Seat counts by route. If Ryanair adds 30% more MCR→IBZ seats for June, that’s a leading indicator. £15–50k/yr.
PHASE 2
ForwardKeys Booking Data — paid, daily, forward-looking
Actual GDS bookings for future travel. Shows booking pace vs same period LY. Balearic tourism board likely already subscribes — partnership route possible.
PHASE 2
Resident Advisor DJ Listings — free, API
Every DJ booking at every club globally. Competitor entertainment tracking. If Amnesia books Calvin Harris the same night, that’s a cannibalisation signal.
PHASE 2
♦ Field Operations · Promoter GPS Tracking · Live
Every promoter, every scan, every £
Track field promoters in real-time via GPS. Every 5 minutes their position updates on the map. Every QR code they hand out is tracked from scan → door entry → EPOS spend. Full attribution: which promoter, which location, which hour generated the most revenue.
17·a
Live promoter shifts
Active Shifts · TodayWho’s on the ground right now
| Promoter | Location | Clock In | Status | GPS Pings | QR Scans | Conversions | Revenue | ROAS |
|---|---|---|---|---|---|---|---|---|
Maria Santos IBIZA AIRPORT | ● IBZ Terminal 1 · Arrivals | 12:00 | ON SHIFT | 48 | 142 | 38 | £4,200 | 6.2x |
Carlos Vega IBIZA AIRPORT | ● IBZ Terminal 1 · Car Rental | 12:00 | ON SHIFT | 48 | 98 | 22 | £2,800 | 4.8x |
Jake Thornton MANCHESTER | ● AO Arena · Pre-show queue | 18:30 | ON SHIFT | 12 | 86 | 0 | £0 | — |
Lisa Morgan NEWCASTLE | ● Newcastle Central Station | 17:00 | ON SHIFT | 24 | 64 | 18 | £1,620 | 5.4x |
Pablo Diaz IBIZA | ● Marina Botafoch · Yacht berths | 14:00 | ON SHIFT | 36 | 12 | 4 | £3,200 | 18.2x |
Sophie Blake LONDON | ● Kings Cross / St Pancras | 16:00 | ON SHIFT | 28 | 54 | 14 | £1,820 | 4.2x |
Ahmad Rashid DUBAI | ● Dubai Marina Walk | 18:00 | ON SHIFT | 8 | 42 | 8 | £1,440 | 3.8x |
Ben Taylor NEWCASTLE | ● Quayside / Neptune House area | 19:00 | ON SHIFT | 6 | 34 | 0 | £0 | — |
Zara Hussain MANCHESTER | Deansgate · Hotel Zone | 11:00 | DONE | 48 | 124 | 32 | £3,840 | 5.8x |
Chris Ward NEWCASTLE | Northumbria Uni Campus | 10:00 | DONE | 56 | 88 | 24 | £1,920 | 4.6x |
17·b
Full attribution chain
How Every £ Is TrackedFrom promoter’s hand to the till
👤
Promoter Scans QR
GPS location captured. Promoter ID + shift ID + timestamp logged. Unique QR code generated per handout.
847
SCANS TODAY
📱
Customer Gets Code
Physical flyer with QR or digital ticket sent via SMS/WhatsApp. Code contains: venue, offer, expiry, promoter attribution.
847
DISTRIBUTED
🚪
Customer Arrives at Door
Door staff scans QR code. System confirms: valid, venue match, not expired, not already redeemed. Entry logged with timestamp.
228
DOOR ENTRIES (26.9%)
🍺
EPOS Spend Tracked
Customer’s QR code linked to tab/card payment. All drinks, food, VIP purchases attributed back to the promoter who handed out the code.
£18.4k
ATTRIBUTED REV
📈
ROI Calculated
Total promoter cost (hours × rate + print) vs attributed revenue. Per-promoter, per-location, per-hour ROAS. Identifies best locations and best staff.
5.8x
AVG ROAS
17·c
Location performance ranking
| Location | Total Scans | Conversions | Conv % | Revenue | Cost | ROAS | Best Hour | Verdict |
|---|---|---|---|---|---|---|---|---|
Marina Botafoch, Ibiza Yacht berths · VIP demographic | 12 | 4 | 33.3% | £3,200 | £176 | 18.2x | 15:00–16:00 | SCALE |
IBZ Airport, Terminal 1 Arrivals hall · UK flights | 240 | 60 | 25.0% | £7,000 | £672 | 10.4x | 14:00–16:00 | SCALE |
Deansgate, Manchester Hotel zone · walking route to venues | 124 | 32 | 25.8% | £3,840 | £336 | 11.4x | 19:00–20:00 | SCALE |
Newcastle Central Station Friday/Saturday arrivals | 64 | 18 | 28.1% | £1,620 | £269 | 6.0x | 18:00–19:00 | HOLD |
Northumbria Uni Campus Student halls · daytime promo | 88 | 24 | 27.3% | £1,920 | £336 | 5.7x | 12:00–13:00 | HOLD |
Kings Cross / St Pancras Eurostar + LNER arrivals | 54 | 14 | 25.9% | £1,820 | £336 | 5.4x | 17:00–18:00 | HOLD |
AO Arena Queue, Manchester Pre-concert queue · captive audience | 86 | 0 | TBD | £0 | £168 | — | — | TONIGHT |
Dubai Marina Walk Evening promenade · tourists | 42 | 8 | 19.0% | £1,440 | £202 | 7.1x | 19:00–20:00 | HOLD |
17·d
GPS trajectory insights
Movement Analysis · TodayWhere did promoters spend their time?
Maria Santos (IBZ Airport) — Optimal coverage
142 scans across 8 hours. Best zone: Gate 3 arrivals (UK flights land here). 68% of scans within 50m of exit doors. Movement pattern: efficient figure-8 between arrivals and car rental. No dead zones. Top performer — recommend for weekend peak shifts.
142
Carlos Vega (IBZ Airport) — Dead zone detected
98 scans but GPS shows 45 minutes spent in airport café area (12:30–13:15). Zero scans during this period. Missed easyJet U28421 from Manchester (189 pax, landed 12:42). AI recommendation: stagger lunch breaks so arrivals hall is always covered.
45m GAP
Jake Thornton (AO Arena) — Pre-show position confirmed
Arrived 18:30, positioned outside main entrance. Queue forming from 19:00. 86 scans in 90 minutes. Conversion rate pending (show hasn’t ended yet — customers will arrive at Factory 251 post-show ~22:30). Projected conversion: 22–28 based on historical Drake show data.
86
Ben Taylor (Newcastle Quayside) — Wrong location
GPS shows Ben walked to Quayside at 19:00 but spent 40 minutes at The Bridge Tavern (not a Tokyo venue). Only 34 scans in 2 hours. Should be at Diamond Strip (Collingwood St / Neville St) where foot traffic is 4x higher at this hour. AI recommendation: reassign to Bigg Market / Collingwood Street corridor. Auto-notification sent to area manager.
REASSIGN
17·e
Promoter leaderboard
| # | Promoter | Shifts | Hours | Total Scans | Conversions | Conv % | Revenue | Cost | ROAS | £/Hour |
|---|---|---|---|---|---|---|---|---|---|---|
| 01 | Maria Santos IBZ Airport · Ibiza | 22 | 176 | 3,124 | 842 | 27.0% | £92,400 | £2,957 | 31.2x | £525 |
| 02 | Pablo Diaz Marina Botafoch · Ibiza | 18 | 108 | 216 | 72 | 33.3% | £57,600 | £1,814 | 31.7x | £533 |
| 03 | Zara Hussain Deansgate · Manchester | 20 | 140 | 2,480 | 640 | 25.8% | £76,800 | £2,352 | 32.7x | £549 |
| 04 | Lisa Morgan Central Station · Newcastle | 16 | 96 | 1,024 | 288 | 28.1% | £25,920 | £1,613 | 16.1x | £270 |
| 05 | Sophie Blake Kings Cross · London | 14 | 84 | 756 | 196 | 25.9% | £25,480 | £1,411 | 18.1x | £303 |
| All promoters (14) | 148 | 984 | 12,400 | 3,348 | 27.0% | £368,200 | £16,531 | 22.3x | £374 |
17·f
Shift scheduling
Next 7 Days · Promoter RosterAI-optimised based on flight data + events
Sun 20 Apr — IBZ Airport: 2 promoters (Maria + Carlos) 12:00–20:00
8 UK flights arriving. Peak: 14:00–16:00 (4 flights in 2 hours = 720 pax). Budget: £672. Projected scans: 280. Projected revenue: £8,400.
CONFIRMED
Tue 22 Apr — AO Arena MCR: 1 promoter (Jake) 18:00–23:00
Biffy Clyro concert. 18,000 capacity. Post-show crowd heads to Deansgate. Position at arena exit from 22:00. Projected conversions: 35. Revenue: £2,800.
CONFIRMED
Sat 26 Apr — St James’ Park NCL: 2 promoters 14:00–18:00
Newcastle vs Chelsea (15:00 KO). 52,000 fans. Post-match flow towards Diamond Strip. Position at Gallowgate + St James’ Boulevard. Target: 180 scans, 50 conversions.
PENDING
AI Suggestion: Add Dubai Marina promoter for Thu brunch nights
Thursday is Dubai’s biggest nightlife night (brunch culture). Currently no promoter scheduled for Thu evenings at JBR/Marina Walk. Data shows 42% of Dubai Mayfair walk-ins on Thu come from Marina Walk direction. Cost: £134/shift. Projected revenue: £2,200. ROAS: 16.4x.
SUGGEST
♦ Reservations & Bookings · Live Integration
Every table, every walk-in
Pulse integrates with DesignMyNight, SevenRooms, Collins, and ResDiary via API to pull real-time reservation data, pre-booked revenue, door entry counts, and table availability across all 68 venues. Pre-booked revenue is the most reliable forward indicator — it feeds directly into the ML prediction model.
13·a
Tonight’s reservations by venue
Table Bookings & Pre-Booked RevenueSaturday 19 April · all venues
| Venue | Bookings | Covers | Pre-Booked £ | Tables Left | Walk-in Est. | Door Count | Platform | Status | |
|---|---|---|---|---|---|---|---|---|---|
| 01 | Impossible MANCHESTER | 84 | 342 | £28,400 | 12 / 96 | +480 | 612 | DMN | FULL |
| 02 | The Emerald Theatre LONDON W1 | 62 | 248 | £24,200 | 8 / 70 | +320 | 428 | 7ROOMS | 88% |
| 03 | One Ninety Four LONDON W1 | 48 | 192 | £19,800 | 4 / 52 | +180 | 284 | 7ROOMS | 92% |
| 04 | Digital NEWCASTLE | 72 | 288 | £18,600 | 18 / 82 | +640 | 748 | DMN | 78% |
| 05 | Impossible YORK | 56 | 224 | £14,200 | 8 / 64 | +280 | 386 | DMN | 87% |
| 06 | Dubai Mayfair DUBAI | 44 | 176 | £12,800 | 6 / 48 | +120 | 224 | 7ROOMS | 87% |
| 07 | Tup Tup Palace NEWCASTLE | 38 | 152 | £9,400 | 14 / 48 | +420 | 512 | DMN | 71% |
| 08 | BLVD MANCHESTER | 32 | 128 | £8,200 | 10 / 42 | +340 | 402 | COLLINS | 76% |
| 09 | Gorilla MANCHESTER | 28 | 112 | £6,800 | 16 / 44 | +280 | 348 | DMN | 64% |
| 10 | Factory (FAC251) MANCHESTER | 22 | 88 | £4,200 | 0 / 28 | +560 | 624 | DMN | FULL |
| 11 | Church LEEDS | 14 | 56 | £2,800 | 22 / 36 | +180 | 198 | DMN | 39% |
| 12 | Tower Ballroom HULL | 6 | 24 | £1,200 | 28 / 32 | +80 | 88 | RESDIARY | 12% |
Total reservations tonight
2,847
+18% vs same Sat LW
Pre-booked revenue
£186k
38% of predicted total £487k
Table utilisation
72%
862 / 1,204 tables filled
Door count · live
4,854
Walk-ins: 3,880 · Booked: 974
13·b
Platform integrations
DesignMyNight
38 venues connected · Primary UK platform
LIVE
API: [REDACTED] · OAuth2 · Webhook push on new booking
Data: Reservations, covers, pre-orders, deposit value, table assignment, special requests, no-show history
Refresh: Real-time webhooks + 15-min poll backup
ML feature: dmn_bookings_tonight, dmn_prebooked_revenue, dmn_covers_vs_avg, dmn_noshow_rate
SevenRooms
8 venues · Premium London + Dubai
LIVE
API: [REDACTED] · API key · REST + webhooks
Data: Guest profiles, CRM tags, spend history, VIP status, auto-tagging, table management, waitlist
Refresh: Real-time webhooks · Guest CRM sync daily
ML feature: sr_vip_bookings, sr_avg_spend_per_cover, sr_repeat_guest_pct, sr_waitlist_length
Collins
12 venues · Bar & restaurant focus
LIVE
API: [REDACTED] · API key · JSON REST
Data: Bookings, covers, pre-orders, deposits, table plan, availability, special occasions
Refresh: Webhooks on booking/cancellation + hourly sync
ML feature: collins_covers_tonight, collins_deposit_total, collins_cancellation_rate
ResDiary
4 venues · Legacy integration
MIGRATING
API: [REDACTED] · API key · XML/JSON
Data: Bookings, covers, availability. Limited CRM. Migrating to DMN.
Refresh: Hourly poll (no webhook support)
Note: Hull, Huddersfield, Durham, Liverpool. Moving to DMN by Q3 2026.
13·c
Door entry & live capacity
Entrance Counters · TonightWho’s in the building right now
Impossible MCR — 612 inside (capacity 800)
Door count: 612 in, 188 out. Net: 424 currently inside. Peak hour 22:00–23:00 (148 entries). Queue: ~40 people, 12 min wait. Booked tables: 84/96 occupied. Walk-in conversion: 94%.
76%
Digital NCL — 748 inside (capacity 1,200)
Peggy Gou night. Advance tickets: 1,240 sold. Door: 748 in so far. Expected peak 01:00 (capacity 92%). VIP area: 28/32 tables booked. Guest list: 180 names, 142 arrived (79%).
62%
Factory MCR — 624 inside (capacity 650)
Four Tet night. NEAR CAPACITY. Door queue: ~80 people, 22 min wait. One-in-one-out from 23:15. Pre-sold tickets: 680 (oversold by 30 — no-show buffer). Agent sent “selling fast” push notification at 21:00.
96%
Tup Tup Palace NCL — 512 inside (capacity 700)
Saturday residency. Walk-ins strong tonight (payday effect). Pre-booked VIP: 38 tables (152 covers). Bottle service pre-orders: £4,800. Queue building from 23:30.
73%
Tower Ballroom Hull — 88 inside (capacity 1,200)
7% capacity. 6 table bookings (24 covers). Walk-ins: 64. No queue. Door staff overstaffed at 4 (need 1). Agent recommends: stand down 3 door staff, save £180. Marketing paused (zero ROI tonight).
7%
13·d
7-day booking pipeline
| Day | Bookings | Covers | Pre-Booked £ | vs LW | Pace | Key Events |
|---|---|---|---|---|---|---|
| Sat 19 (today) | 2,847 | 11,388 | £186,400 | +18% | STRONG | Peggy Gou, Four Tet, payday |
| Sun 20 | 420 | 1,680 | £28,200 | +8% | NORMAL | Sunday roasts, brunches |
| Mon 21 | 180 | 720 | £12,400 | 0% | QUIET | Industry night @ Digital |
| Tue 22 | 210 | 840 | £14,800 | −4% | QUIET | Quiz nights x 8 venues |
| Wed 23 | 340 | 1,360 | £22,600 | +12% | BUILDING | Student night @ 6 venues |
| Thu 24 | 620 | 2,480 | £42,800 | +22% | STRONG | Pay day eve, corporate bookings |
| Fri 25 | 1,840 | 7,360 | £124,600 | +28% | PEAK | Payday Friday, Drake after-party |
| Sat 26 | 2,210 | 8,840 | £162,200 | +14% | STRONG | Payday Sat, Sonny Fodera @ Digital |
Example Predictions · Static Demo Data
♦ Master Report · Group Consolidated Intelligence
Every metric, every venue, one report
The Master Report consolidates every quantifiable metric across all 68 venues, 7 regions, and 4 continents into a single actionable document. Every number is reasoned by our AI agent panel. Every problem has a proposed solution. Every win is quantified and attributed.
18·a
Executive summary
Group Revenue (Wk)
£2.84M
+8.4% vs LW
Group EBITDA
£428k
15.1% margin
Avg GP %
71.2%
+0.8pp vs target
Labour %
23.4%
−0.6pp vs plan
Marketing ROAS
2.4x
vs 1.8x LW
Venues Above BE
54/68
79.4%
Open Issues
12
3 critical
Actions Overdue
4
escalated
18·b
Regional breakdown
| Region | Venues | Revenue | vs LW | GP % | Labour % | EBITDA | Mktg ROAS | Issues | Health |
|---|---|---|---|---|---|---|---|---|---|
| North East | 7 | £624k | +12.1% | 71.8% | 23.2% | £94k | 2.6x | 3 | STRONG |
| North West | 8 | £742k | +8.4% | 72.1% | 22.8% | £118k | 2.2x | 1 | STRONG |
| Yorkshire | 6 | £268k | −4.2% | 67.8% | 28.6% | £32k | 1.4x | 5 | REVIEW |
| London | 4 | £412k | +14.2% | 69.4% | 25.6% | £62k | 2.8x | 1 | STRONG |
| Balearics | 4 | £384k | +22.4% | 74.6% | 18.4% | £82k | 3.2x | 0 | PEAK |
| Middle East | 2 | £202k | +8.8% | 72.1% | 21.2% | £38k | 2.1x | 0 | STABLE |
| USA | 4 | £168k | +1.2% | 65.8% | 27.8% | £18k | 1.6x | 2 | WATCH |
18·c
AI-reasoned problems & solutions
Every issue identified, reasoned, and solvedThis week’s action list
Yorkshire region: Labour at 28.6% (target 24%) — £14.2k weekly overspend
Problem: Church Leeds (31.4%), Tower Hull (34.2%), BrewDog Huddersfield (32.4%) all overstaffed. Revenue doesn’t support headcount.
Solution (Staffing Agent): Reduce Church Sat rota 14→11 bar, Tower 12→8 bar + 8→4 door, BrewDog close 1hr earlier. Combined: £6,160/wk saving, £320k/yr.
Owner: Sarah Collins (Yorkshire Area) · Due: 22 Apr · Status: PENDING
Solution (Staffing Agent): Reduce Church Sat rota 14→11 bar, Tower 12→8 bar + 8→4 door, BrewDog close 1hr earlier. Combined: £6,160/wk saving, £320k/yr.
Owner: Sarah Collins (Yorkshire Area) · Due: 22 Apr · Status: PENDING
£320k/yr
Tower Hull: Systematic shrinkage — 380 voids + £300 cash variance in 60 days
Problem: Cross-correlated voids + cash variance at Terminal 3, Thu/Sat 22:00–01:00. Pattern indicates staff-level issue.
Solution (Loss Prevention Agent): Immediate GM investigation. CCTV review of Terminal 3 for flagged sessions. Reassign till operator. Install pour monitors on 3 highest-void lines.
Owner: Dave Mitchell (GM) · Due: 21 Apr · Status: OVERDUE (day 9)
Solution (Loss Prevention Agent): Immediate GM investigation. CCTV review of Terminal 3 for flagged sessions. Reassign till operator. Install pour monitors on 3 highest-void lines.
Owner: Dave Mitchell (GM) · Due: 21 Apr · Status: OVERDUE (day 9)
£54k/yr
Group burn rate: £842k annual saving available from optimal trading hours
Problem: 49 UK venues collectively burn £16.2k/week on early-open and late-close hours where revenue < cost.
Solution (Operations Agent): Implement venue-by-venue optimal hours. Phase 1 (5 venues) approved — Impossible MCR already saving £480/Sat. Phase 2: roll to 20 more venues by May.
Owner: James Wright (Ops Director) · Timeline: 4 weeks
Solution (Operations Agent): Implement venue-by-venue optimal hours. Phase 1 (5 venues) approved — Impossible MCR already saving £480/Sat. Phase 2: roll to 20 more venues by May.
Owner: James Wright (Ops Director) · Timeline: 4 weeks
£842k/yr
Champagne promo timing: £93.6k annual revenue uplift available
Problem: 8 venues do first VIP sparkler show at 00:30. Data shows 23:30 generates 42% more follow-on purchases.
Solution (Revenue Agent): Digital NCL moved to 23:30 — +28% champagne rev confirmed. Roll to remaining 7 VIP-table venues.
Status: 1/8 implemented. 7 pending deployment.
Solution (Revenue Agent): Digital NCL moved to 23:30 — +28% champagne rev confirmed. Roll to remaining 7 VIP-table venues.
Status: 1/8 implemented. 7 pending deployment.
£93.6k/yr
Balearics region: Best-performing region — 74.6% GP, 18.4% labour, +22.4% WoW
Why: Ibiza season opening + low labour (seasonal staff model) + high avg spend (€85/hd). AMOK Mallorca highest GP in group (75.2%). Hangar 8289 benefiting from UK payday weekend flight surge.
Recommendation (Growth Agent): Scale Ibiza marketing budget +40% for May. Secure additional weekend airport promoter shifts. Pre-book DJ lineup for June opening parties.
Recommendation (Growth Agent): Scale Ibiza marketing budget +40% for May. Secure additional weekend airport promoter shifts. Pre-book DJ lineup for June opening parties.
WIN
Field Ops promoters generating 22.3x ROAS — £368k attributed in 30 days
Data: 14 promoters, 984 hours, 12,400 QR scans, 3,348 door conversions. Best: Marina Botafoch yacht berths (31.7x ROAS).
Recommendation (Growth Agent): Add 4 more promoters: Dubai Marina Thu (16.4x projected), Leeds station Fri, Manchester Arndale Sat, Newcastle Quayside repositioned to Diamond Strip.
Recommendation (Growth Agent): Add 4 more promoters: Dubai Marina Thu (16.4x projected), Leeds station Fri, Manchester Arndale Sat, Newcastle Quayside repositioned to Diamond Strip.
WIN
18·d
What to do next week
AI Consolidated RecommendationsTop 5 actions to make next Saturday better
Resolve Tower Hull voids immediately — estimated £54k/yr leakage
Day 9 with no action. Escalate to Aaron on Tuesday call. GM must produce CCTV review + action plan by Wednesday. If unresolved by Friday, trigger mystery shopper + till reassignment.
Roll optimal trading hours to 20 more venues — £520k potential annual saving
Impossible MCR proof-of-concept confirmed (£480/Sat saving). Next: Factory, BLVD, Gorilla (MCR cluster), then Digital, Tup Tup, Neptune (NCL cluster). Area managers to submit current vs proposed hours by Friday.
Pre-position for Klute Durham reading week (starts 28 Apr)
Cut Klute marketing to zero from 25 Apr. Reduce Saturday staff by 4. Prepare “welcome back” campaign for 5 May return. Redirect Klute ad budget to Digital NCL for same period.
Roll champagne promo timing to 7 remaining VIP venues
Digital NCL confirmed +28% uplift. Next: Impossible MCR, Tup Tup, BLVD, Emerald Theatre, One Ninety Four, Dubai Mayfair, Hangar 8289. GM briefing document auto-generated — send Monday AM.
Payday Saturday (26 Apr): boost marketing +30%, extend all trading hours +45 min
Historical data: payday Saturdays generate +18% group revenue. Model predicts £575k (vs £487k this week). Pre-book extra staff for 22:00–02:00 peak. Activate all arrival triggers at maximum budget.
♦ AI Agent Panel · Specialist Hospitality Intelligence
12 expert agents, one mission
Each agent is a specialist AI model fine-tuned for a specific domain of hospitality operations. They run continuously, analysing data streams, identifying problems before humans see them, and proposing solutions grounded in industry best practices and your group’s historical data.
19·a
Agent roster
🤖 Revenue Optimiser
Pricing · upsell triggers · menu engineering
LIVE
Focus: Avg spend per head, drink pricing elasticity, VIP package optimisation, champagne trigger timing, menu item profitability ranking. This week: Identified £93.6k champagne uplift from sparkler timing. Recommended £0.50 price increase on 4 spirit lines (margin gain £42k/yr, zero volume impact predicted).
💰 Cost Controller
Labour · COGS · utilities · waste
LIVE
Focus: Labour % by venue/hour, optimal trading hours, staff rota efficiency, supplier cost tracking, energy consumption. This week: Flagged £842k annual saving from optimal hours. Detected Church Leeds over-rostered by 3 staff. Identified £2.4k/wk energy saving from AC scheduling at Dubai venues.
🔎 Loss Prevention
Voids · cash variance · shrinkage · theft
LIVE
Focus: Void patterns (operator × terminal × time), cash-up discrepancies, stock variance vs POS, CCTV cross-referencing. This week: Tower Hull escalated to critical. Klute Durham cash variance £240 flagged. City Vaults £140 under investigation. Group shrinkage down from 4.2% to 2.8% since agent deployment.
🎯 Marketing Strategist
Campaigns · ROAS · creative · audience
LIVE
Focus: Campaign performance by channel, creative A/B testing, audience targeting, budget allocation, arrival trigger campaigns. This week: Reallocated £800 from Impossible MCR (diminishing returns) to Tower Hull rescue campaign. Concert exit campaigns achieving 12.9x ROAS. Paused Church Leeds spend (ops must fix first).
👥 Customer Retention
Loyalty · repeat visits · win-back · NPS
LIVE
Focus: Repeat visit frequency, lapsed customer win-back, birthday/anniversary offers, VIP loyalty tiers, post-visit surveys. This week: Identified 2,400 lapsed VIPs (no visit in 60+ days). Auto-generated personalised win-back offers via WhatsApp. 18% open rate, 4.2% rebooking rate. £12.8k recovered revenue.
🎓 Student Specialist
Term dates · Freshers · exam seasons · loans
LIVE
Focus: University calendar tracking (6 institutions), student loan drop dates, Freshers campaign planning, exam season budget cuts, graduation premium pricing. This week: Pre-alerted Klute Durham reading week (28 Apr). Freshers 2026 campaign brief auto-generated for September. Student loan Apr drop effect now in decay phase.
✈ Arrival Trigger Agent
Flights · events · geo-fencing · field ops
LIVE
Focus: 30 trigger types across transport, events, “getting ready” signals. Auto-deploys campaigns within 5 mins of trigger. Manages promoter scheduling. This week: 2,840 triggers fired. 568 campaigns auto-deployed. Group trigger ROAS: 9.1x. Top: weather triggers 28.3x, Snapchat geofilters 14.1x.
☀ Weather Analyst
Forecasts · terrace ops · rota impact
LIVE
Focus: 5-day weather per venue, terrace threshold (>14°C), rain impact model, storm warning budget cuts, optimal staffing adjustments. This week: Dry 16°C Saturday triggered terrace campaigns at 5 venues. Pre-booked extra terrace staff at Impossible MCR. Saved £4.2k by pausing ads before Thu rain.
📊 Forecasting Engine
Predictions · baselines · attribution
LIVE
Focus: 15-day revenue prediction per venue, baseline vs marketing uplift attribution, 12-week Saturday forecast, regional heat scoring. This week: 96.9% accuracy maintained. Predicted Sat 19 Apr at £487k (actual £487.3k). Next Sat forecast: £575k (payday + dry weather).
🍺 Drinks Intelligence
Category mix · pour accuracy · trends
LIVE
Focus: Hourly drink category shifts, champagne trigger correlation, spirit-to-cocktail ratio as crowd predictor, pour vs POS variance, stock re-order optimisation. This week: Detected over-pouring at 3 venues (Grey Goose measure +12% vs POS). Recommended spirit-to-cocktail ratio monitoring as close-time predictor across group.
📜 Compliance Monitor
Licensing · H&S · CCTV · training
LIVE
Focus: License renewal tracking, fire safety cert expiry, DBS check monitoring for door staff, CCTV retention compliance, staff training completion rates. This week: Flagged 3 door staff DBS renewals due within 30 days. Tower Hull fire cert inspection due 28 Apr. All venues: CCTV retention compliant.
💡 Innovation Scout
Competitors · trends · new concepts · tech
LIVE
Focus: Competitor monitoring (pricing, events, social), industry trend analysis, new concept evaluation, technology scouting. This week: Warehouse Project announced summer residency — 18% demo overlap with Factory. Recommended counter-programming with exclusive DJ bookings on same dates. Identified TikTok cocktail trend (+2.1M views) that Impossible MCR can replicate.
♦ Arrival Intelligence · Moment-Triggered Campaign Engine
Right person, right place, right second
30 trigger types detect when potential customers arrive at known locations — airports, arenas, hotels, stations, gyms, salons — and automatically deploy targeted ad campaigns within minutes. Digital triggers execute autonomously via our 55-agent marketing engine. Physical activations assign on-site staff with tracked ROI.
16·a
Live trigger feed
FLIGHT LANDED: FR1842 Manchester → Ibiza · 189 pax · 72% aged 18–34
Auto-deployed: Instagram Reels + TikTok ads targeting IBZ airport 2-mile radius. Creative: “Just landed? Hangar 8289 tonight. Free entry before midnight.” Budget: £45 for 3 hours. Agent: flight-arrival-ibz-042
14:22
CONCERT EXIT: Drake @ AO Arena Manchester · 21,000 capacity · Ends ~22:30
Pre-scheduled: Instagram Stories + Snapchat geofilter activated at 22:15. 500m radius around arena. Creative: “Show’s over but the night isn’t. Factory 251 — 3 min walk.” Budget: £120. Agent: arena-exit-mcr-019
PRE-SET
WEATHER TRIGGER: Manchester 19°C & dry Saturday · Terrace campaigns activated
WeatherAds.io fired at 08:00. Auto-activated Impossible MCR terrace campaign on Meta + Google. Creative: “Sun’s out. Rooftop open from 2pm.” Budget: £60/day. Agent: weather-terrace-mcr-008
08:00
BEAUTY EXIT: 14 nail bars & salons in Deansgate area · Saturday 14:00–18:00
Geo-fence active. Instagram Stories targeting women 18–35 leaving grooming businesses. Creative: “Fresh nails deserve a fresh night out. Impossible MCR tonight.” Budget: £35. Agent: beauty-exit-mcr-003
14:00
YACHT ARRIVAL: 3 superyachts docked Marina Botafoch Ibiza · ~36 high-value guests
Tight geo-fence (500m). Instagram + WhatsApp Business. Creative: “Docked at Botafoch? VIP table reserved. WhatsApp us.” Budget: £25. Avg table spend: £2,800. Agent: yacht-ibz-007
13:45
PHYSICAL: Airport welcome team — Ibiza · 2 staff assigned · Terminal arrivals hall
Staff: Maria Santos + Carlos Vega. Shift: 12:00–20:00. Task: distribute Hangar 8289 flyers + guestlist QR codes to arriving UK flights. ROI tracked via unique QR scan count. Scans today: 142 · Estimated conversions: 38
ON-SITE
16·b
30 trigger types
Transport Arrivals · 8 triggers
✈ Flight Arrivals
IBZ · PMI · DXB · LAX · JFK · LHR
AUTO
Trigger: Flight API webhook on landing. Radius: 3km airport. Window: 3hrs post-landing. Platform: Instagram Reels + TikTok. Budget: £45–100/flight. CTR: 5–8%. ROAS: 9/10
🚆 Train Arrivals
Piccadilly · Central · Kings Cross · Leeds City
AUTO
Trigger: Fri/Sat 17:00–22:00 schedule. Radius: 1km station. Platform: Instagram Stories + Snapchat. Budget: £30–50/station/weekend. ROAS: 7/10
🚢 Cruise Ship Docking
Ibiza port · Hvar harbour · Dubai Harbour
AUTO
Trigger: MarineTraffic AIS alert. Audience: 3,000–6,000 pax per ship. Budget: £20–50/port call. ROAS: 6/10
⛵ Yacht Marina
Marina Botafoch · Hvar · Dubai Marina · Marina del Rey
AUTO
Trigger: Tight 500m geo-fence. Audience: 6–12 per yacht, ultra-high-value. Avg spend: £500–5,000. Platform: Instagram + WhatsApp. ROAS: 8/10
🚌 Ferry Arrivals
Split→Hvar · Barcelona→Ibiza · Denia→Ibiza
AUTO
Trigger: Ferry schedule + geo-fence terminal. Audience: 500–1,500/ferry. Budget: £15–30/day summer. ROAS: 6/10
🚕 Coach Arrivals
Megabus · National Express · FlixBus
AUTO
Trigger: Fri/Sat 16:00–21:00. Demo: Budget 18–25 yr olds. Creative: “Free entry tonight — show this ad.” ROAS: 6/10
🚘 Rental Car Pickups
IBZ · PMI · DXB · LAX airport car lots
AUTO
Trigger: Geo-fence rental car lot (separate from terminal). Audience: Multi-day visitors. ROAS: 5/10
🚃 Eurostar Arrivals
St Pancras · Paris/Brussels/Amsterdam arrivals
AUTO
Trigger: Weekend arrivals 14:00–20:00. Language: Multi-language creative (FR/NL/DE). Demo: Affluent European visitors. ROAS: 7/10
Event Exits · 4 triggers
🎵 Concert & Arena Exits
AO Arena · Utilita · First Direct · O2
AUTO
Trigger: Show end time from Ticketmaster API. Audience: 15,000–21,000 per show. Radius: 500m. CTR: 5–8% (3–4x normal). Budget: £50–150/event. ROAS: 10/10. Factory 251 is a 3-min walk from AO Arena.
⚽ Sports Match Exits
Old Trafford · Etihad · St James’ Park · Elland Road
AUTO
Trigger: Full-time from live scores API. Audience: 37,000–75,000. Dynamic creative: Win = “Celebrating?” / Loss = “Drown your sorrows.” ROAS: 8/10
🎪 Festival Exits
Parklife · Creamfields · Reading/Leeds · Wireless
AUTO
Trigger: Final day geo-fence + 48hr device ID retarget. Audience: 70,000–210,000. Budget: £200–500/festival. ROAS: 9/10. Parklife returns 80k to Manchester — afterparty pipeline.
💼 Conference Exits
Manchester Central · ExCeL · NEC · Harrogate
AUTO
Trigger: Conference closing time 16:00–19:00. Audience: 5,000–20,000 delegates on expenses. Avg spend: £68/head (1.6x regular). ROAS: 8/10
“Getting Ready” Signals · 5 triggers
💇 Beauty & Grooming Exits
Hairdressers · nail bars · barbers · tanning
AUTO
Signal strength: 90%+ going-out intent. Someone who spent £100 on hair/nails is going out tonight. Geo-fence: Beauty business clusters on high streets. Fri 12–19:00, Sat 10–18:00. CTR: 4–6%. ROAS: 8/10
🏋 Gym Exits
PureGym · JD Gyms · The Gym Group · city centres
AUTO
Trigger: Fri 14:00–19:00, Sat 12:00–18:00. Working out before going out. Creative: “Post-gym glow? Show it off at [Venue] tonight.” ROAS: 7/10
🛍 Fashion Shopping
Arndale · Trinity · Eldon Square · Oxford Street
AUTO
Trigger: Fri/Sat 12:00–18:00 in fashion retail zones. Creative: “New outfit deserves a new venue.” Audience: 41M+ visits/yr at Manchester Arndale alone. ROAS: 6/10
🍺 Pre-Drink Purchases
Supermarkets · off-licences · Fri/Sat 17:00–21:00
AUTO
Trigger: Geo-fence supermarkets within 3mi of venues. Signal: Buying alcohol on Friday evening = going out. Creative: “Pre-drinks sorted? Now sort the main event.” ROAS: 7/10
🚗 Ride-Hailing Orders
Uber Journey Ads · in-app + in-car tablet
AUTO
Platform: Uber Advertising. Reach: 9/10 leisure riders notice ads. Format: In-app display + in-car tablet video. Creative: “On your way out? Make it to [Venue].” ROAS: 7/10
Accommodation · 2 triggers
🏨 Hotel Check-In Zones
Playa d’en Bossa · JBR Dubai · Deansgate MCR
AUTO
Trigger: Meta “People travelling in this location” within hotel cluster. Window: Check-in 14:00–18:00, then pre-going-out 19:00–22:00. CTR: 4–6%. ROAS: 8/10
🏠 Airbnb/Short-Let Zones
Party areas · Northern Quarter · San Antonio
AUTO
Trigger: Same as hotels but residential zones with known short-let density. Meta filters out residents with “travelling in” targeting. ROAS: 7/10
Seasonal & Life Events · 7 triggers
🎓 Freshers Week
Sep/Oct · 50,000+ students arriving per city
AUTO
Why 10/10: Capture a student in week 1 = 3 years of regular visits. Geo-fence: Campuses + student accommodation (Unite, Liberty Living). Budget: £500–1,000/city for 2-week blitz. Platform: TikTok + Snapchat (primary for 18–21). ROAS: 10/10
💍 Stag & Hen Exits
Go-karting · escape rooms · paintball · cocktail classes
AUTO
Trigger: Geo-fence activity venues Sat 14:00–18:00. Creative: “Activity done? VIP reserved. Group discounts for 10+.” Avg group spend: £500–2,000. ROAS: 9/10
☀ Weather Triggers
Temp >18°C + dry = terrace campaigns
AUTO
Tool: WeatherAds.io (£49/mo). CTR lift: +80.7% vs non-weather-targeted. Automation: Rule fires → Meta + Google campaigns auto-activate. Zero manual intervention. ROAS: 8/10
💰 Payday Weekends
Last Fri of month + first weekend
AUTO
Trigger: Calendar-scheduled. Boost: +30% budget on payday weekends. Creative: “Payday weekend. Treat yourself. VIP at [Venue].” ROAS: 7/10
💗 Dating App Proximity
City centre Fri/Sat evening swipers
AUTO
Targeting: Meta interests: dating apps + geo + Fri/Sat evening. Creative: “Need a first date spot? [Venue] — impressive without trying too hard.” ROAS: 6/10
🎓 Graduation Ceremonies
Jul graduation weeks · families + graduates
AUTO
Trigger: Geo-fence ceremony venues. Audience: 3,000+ per ceremony. Creative: “Graduated? Celebrate at [Venue]. Group bookings open.” Avg spend: 2.4x normal (family premium). ROAS: 7/10
📷 Snapchat Geofilters
Always-on at every venue · £5–20/night
AUTO
Cost: As low as £5/night per venue. Reach: Every photo/video with filter = free advertising to all their friends. Benchmark: 22,000 impressions for £35 (12% conversion). ROAS: 9/10. Deploy across all 68 venues permanently.
Physical Activations · 4 triggers (staff-assigned + ROI tracked)
👤 Airport Welcome Teams
IBZ · PMI · DXB arrivals hall
STAFF
Assignment: 2 staff per airport, peak arrival windows. Task: Flyers + guestlist QR codes to arriving UK flights. ROI tracking: Unique QR scan → guestlist signup → door entry → EPOS spend. Full attribution chain. Cost: £16.80/hr/staff + print. Measured ROAS: 4–6x
🚗 Branded Limo/Taxi
Airport transfers with venue branding
STAFF
Assignment: Partner with local transfer companies. Integration: Riders receive WhatsApp message with venue info during transfer. ROI tracking: Promo code per transfer company. Measured via: Code redemptions at EPOS.
🏨 Hotel Concierge Partners
Lobby cards · concierge commission · hotel bar promos
STAFF
Assignment: Venue promoter visits hotel weekly, refreshes promotional material. Commission: £5 per referred booking to concierge. ROI tracking: Concierge booking code → reservation platform → EPOS spend. Staff: 1 promoter per destination city.
🎫 Stag/Hen Activity Partners
Cross-promo with go-karting, escape rooms, etc.
STAFF
Assignment: Venue promoter establishes partnerships with 10–15 activity providers per city. Deal: “Show your activity receipt for a free round.” ROI tracking: Receipt scans at door → EPOS spend attributed. Cost: Free round (£4.20) → avg group spend £800. ROI: 190x.
16·c
Campaign performance by trigger
| Trigger Type | Triggers/Wk | Reach | CTR | Conversions | Spend | Revenue | ROAS | Status |
|---|---|---|---|---|---|---|---|---|
| Concert/Arena exits | 4 | 84,000 | 6.2% | 1,240 | £480 | £6,200 | 12.9x | LIVE |
| Snapchat geofilters | 476 | 142,000 | 12.4% | 4,200 | £340 | £4,800 | 14.1x | LIVE |
| Flight arrivals (Ibiza) | 42 | 8,400 | 7.1% | 420 | £1,890 | £14,200 | 7.5x | LIVE |
| Weather triggers | 2 | 220,000 | 1.8% | 880 | £120 | £3,400 | 28.3x | LIVE |
| Beauty/grooming exits | 14 | 4,200 | 4.8% | 142 | £280 | £2,100 | 7.5x | LIVE |
| Hotel check-in zones | 7 | 12,000 | 5.2% | 380 | £420 | £4,800 | 11.4x | LIVE |
| Train station arrivals | 10 | 18,000 | 3.4% | 220 | £300 | £1,800 | 6.0x | LIVE |
| Airport welcome teams | 7 | 2,400 | 18.2% | 142 | £840 | £4,200 | 5.0x | STAFF |
| Stag/hen activity exits | 6 | 480 | 8.4% | 28 | £180 | £2,800 | 15.6x | PHASE 2 |
| Total (all triggers) | 568 | 491,480 | 5.4% | 7,652 | £4,850 | £44,300 | 9.1x | GROUP |
Annual trigger budget
£135k
Projected revenue
£970k
Blended ROAS
7.2x
Triggers per week
2,840
♦ Data Feed Architecture · Technical Breakdown
Every data source explained
Pulse ingests data from 38 distinct feeds across 7 categories. Each feed is classified by cost, update frequency, integration method, and ML feature importance. This page documents every API endpoint, data format, refresh cadence, and how each feed connects to the 70+ factor prediction model.
12·a
Weather & environment feeds
☁
Met Office DataPoint API
LIVE
FREE
Endpoint: [REDACTED] · Met Office DataPoint REST API
Auth: API key (free registration)
Update frequency: Every 3 hours for 5-day forecast · Hourly for observations
Data format: JSON · Temperature (°C), wind speed (mph), wind gust, weather type code, visibility, humidity, UV index, precipitation probability (%)
Coverage: 5,000+ UK locations · We query 30 postcodes (one per venue catchment area)
ML features extracted: temp_celsius, rain_prob_pct, wind_mph, feels_like, uv_index, sunshine_hours, is_dry_saturday (boolean), temp_above_14 (terrace threshold)
Feature importance: 0.14 (4th highest in model). Temperature alone explains 8–12% of Saturday revenue variance.
Rate limit: 5,000 requests/day (we use ~200/day)
Fallback: OpenWeatherMap API (free tier: 1,000 calls/day) if Met Office is down
☀
Open-Meteo API
LIVE
FREE
Endpoint: [REDACTED] · Global weather forecast API
Auth: None required (open API)
Update frequency: Hourly · 16-day forecast · Historical data back to 1940
Coverage: Global — covers Ibiza, Dubai, Hvar, LA, NYC, Palm Springs (Met Office is UK only)
Why both? Met Office is more accurate for UK (national weather service). Open-Meteo covers international venues. Cross-validated for UK locations.
ML features extracted: Same as Met Office + apparent_temperature, direct_radiation (UV proxy), soil_temperature (outdoor ground comfort)
Historical training: 5 years of daily weather × daily EPOS revenue = core training pair for weather features
⚠
Met Office Warnings RSS
LIVE
FREE
Endpoint: [REDACTED] · Severe weather warnings RSS
Data: Amber/Red weather warnings · Storm names · Flood alerts
Action: Amber warning = auto-scale marketing down 48hrs before. Red = suspend all paid campaigns. Saves £4–8k per storm event.
ML feature: severe_weather_active (boolean) · Historically: amber = −22% avg, red = −38% avg
12·b
Aviation & tourism feeds
✈
Eurostat Aviation API
LIVE
FREE
Endpoint: [REDACTED] · EU aviation statistics API
Auth: None (open data)
Update: Monthly (6–8 week lag) · 10+ years historical
Data: Passenger counts by route pair (e.g., MAN→IBZ). All EU/EEA airports. Breakdown by nationality where available.
ML use: Backbone for seasonal flow models. Historical patterns train the drain/gain predictions. Feature: pax_route_monthly, pax_yoy_change
Limitation: Monthly granularity + lag means this is training data only, not real-time. Combined with Google Trends for forward signal.
📈
Google Trends API (pytrends)
LIVE
FREE
Library: [REDACTED] · Search trend analysis tool
Queries tracked: “flights to Ibiza”, “flights to Ibiza from Manchester”, “Ibiza clubs 2026”, “bars near me Manchester”, “clubs tonight Newcastle”, “best nightlife [city]” + 40 more
Geo filter: By city/region (Manchester, Newcastle, Leeds, London)
Update: Weekly aggregate · Daily for high-volume terms
ML use: 4–8 week leading indicator for pax flows. Correlation with Eurostat: 0.82. Free proxy for ForwardKeys (£50k/yr). Feature: search_volume_flights_ibz, search_volume_going_out_mcr
Rate limit: Aggressive — max ~10 requests/min before throttling. We batch overnight.
🏝
Ibestat Tourist Profiles
LIVE
FREE
Source: [REDACTED] · Balearic Islands tourism authority
Data: Tourist arrivals by age (16–24, 25–34, 35–44), nationality, avg daily spend, activities, accommodation type, length of stay
Update: Quarterly · Ibiza + Mallorca specific
ML use: Demographic profiling for destination venues. Tells us exactly what % of Ibiza tourists are in our nightclub target demo. Feature: tourist_pct_18_34_ibz, tourist_avg_spend_gbp
Gold source for: Hangar 8289, B12, AMOK Mallorca revenue modelling. Without this, we’d be guessing demographic mix.
🛫
OAG Scheduled Capacity
PHASE 2
£15–50k/yr
Endpoint: [REDACTED] · Scheduled airline capacity API
Data: Every flight in every airline schedule globally. Seat counts by route. Equipment type (A321neo vs 737-800 = capacity difference). 12 months forward.
ML use: Leading indicator. If Ryanair adds 30% more MCR→IBZ seats for June, that’s a drain prediction 3 months ahead. Feature: scheduled_seats_route_fwd, capacity_yoy_change
Why Phase 2: Expensive. Google Trends + Eurostat covers 80% of the signal at zero cost. OAG adds the remaining 20% precision for £15–50k/yr.
📋
ForwardKeys GDS Bookings
PHASE 2
£40–80k/yr
Source: [REDACTED] · GDS flight booking data provider
Data: Actual GDS bookings for future travel. Booking pace vs same period LY. Origin/destination pairs with booking date and travel date.
ML use: The ultimate forward-looking flight demand signal. Shows actual paid bookings, not just searches. Feature: bookings_pace_vs_ly_pct
Partnership route: Balearic tourism board likely already subscribes. Possible data-sharing agreement via hospitality industry body.
🚢
MarineTraffic AIS API
PHASE 2
£2–8k/yr
Endpoint: [REDACTED] · Maritime vessel tracking API
Data: Cruise ship arrivals (pax count, port call times), yacht movements, ferry schedules
ML use: Cruise ship docking at Ibiza = +200–400 nightlife visitors. Hvar yacht arrivals = premium spend €120/head. Feature: cruise_pax_arriving, yacht_count_weekly
Coverage: Ibiza, Mallorca, Hvar, Dubai · UK ports not relevant for nightlife
💰
Skyscanner Flights API
PHASE 2
FREE (affiliate)
Endpoint: [REDACTED] · Flight pricing & demand API
Data: Flight pricing by route, date range, cheapest month. Price elasticity signal — when MCR→IBZ drops below £50 return, drain increases 40%.
ML use: Price-to-demand elasticity. Feature: flight_price_avg_gbp, flight_price_vs_30d_avg. When flights are cheap, more people leave.
Access: Free via affiliate programme. Requires application + approval (2–4 weeks).
🏛
Dubai Tourism (DTCM) Open Data
PHASE 2
FREE
Source: [REDACTED] · Dubai tourism authority open data
Data: Monthly tourist arrivals by nationality, hotel occupancy, RevPAR, visitor spend. Ramadan dates and public holiday calendar.
ML use: Dubai Mayfair and DIFC modelling. Feature: dubai_occupancy_pct, dubai_uk_arrivals, ramadan_active (boolean). Occupancy >90% = nightlife premium pricing opportunity.
12·c
Economic & macro feeds
🏧
Bank of England API
Endpoint: [REDACTED] · Central bank rates API ·
Data: Base rate, MPC minutes, inflation expectations, money supply ·
Features: boe_base_rate, rate_change_30d, mortgage_payment_delta_monthly ·
Update: 8 meetings/yr + daily yield curves
LIVE
📊
ONS API (Office for National Statistics)
Endpoint: [REDACTED] · National statistics API ·
Series: CPI (D7G7), employment (LF24), wages (KAB9), retail sales (J5EK), claimant count ·
Features: cpi_annual_pct, unemployment_regional, real_wage_growth, retail_sales_mom ·
Update: Monthly · Regional breakdowns for NE, NW, Yorkshire, London
LIVE
💰
GfK Consumer Confidence (via ONS)
Source: GfK survey, published via ONS/press release ·
Data: Overall index + 5 sub-indices (general economy, personal finances, major purchases, savings, expectations) ·
Feature: gfk_index, gfk_major_purchase_intent ·
Update: Last Friday of each month · Ingested within 2 hours of release
LIVE
💱
Exchange Rates API
Endpoint: [REDACTED] · FX rates API ·
Pairs: GBP/EUR, GBP/USD, GBP/AED, GBP/HRK ·
Feature: gbp_eur_rate, gbp_eur_30d_trend ·
Update: Daily · Free tier: 1,500 requests/month
LIVE
⛽
RAC Fuel Watch / BEIS Energy
Sources: [REDACTED] · Fuel pricing & energy regulatory data ·
Features: petrol_ppl, energy_bill_quarterly, taxi_cost_proxy ·
ML use: Cheap taxis = easier late-night transport = +footfall. Energy costs = venue operating margin.
LIVE
🏠
Rightmove / Land Registry HPI
Sources: Land Registry open data + Rightmove monthly index ·
Feature: house_price_yoy_pct, housing_wealth_effect ·
ML use: Wealth effect proxy. Homeowners feel richer → spend more. Low importance (0.02) but statistically significant over 3+ years.
LIVE
💳
HMRC Alcohol Duty Bulletins
Source: [REDACTED] · Government alcohol duty bulletins ·
Data: Duty rates, duty changes, draught relief rates ·
ML use: Feature: duty_frozen_months. Duty freeze = stable pricing = no demand shock. Any change announcement = −2% sentiment dip.
LIVE
12·d
Events & entertainment feeds
⚽
Football-Data.org API
Endpoint: [REDACTED] · Football fixtures API ·
Data: All PL + Championship fixtures, KO times, home/away, stadium location, TV coverage ·
Features: football_match_today (boolean), home_away, ko_time, expected_attendance, tv_coverage_type ·
ML impact: NUFC away = −4% Newcastle venues. Man Utd home late KO = +8% MCR post-match.
LIVE
🎶
Resident Advisor API
Endpoint: [REDACTED] · DJ booking & event listings API ·
Data: Every DJ booking at every club globally. Artist profiles, follower counts, genre tags ·
Features: dj_tier (A/B/C/resident), dj_instagram_followers, genre_match_score, competitor_event_same_night ·
ML impact: A-list DJ = +40–60% uplift. Feature importance: 0.18 (2nd highest in model).
PHASE 2
🎟
Ticketmaster Discovery API
Endpoint: [REDACTED] · Arena & stadium events API ·
Data: Arena/stadium events within 5 miles of each venue. Artist, capacity, start/end times, ticket availability ·
Features: arena_event_tonight (boolean), arena_capacity, event_end_time, genre_overlap_pct ·
ML use: Concert ending at 22:30 = post-show spillover to nearby venues. Drake @ AO Arena = +£8–12k MCR venues.
LIVE
🎫
FIXR / Skiddle / Dice / Fatsoma APIs
Data: Advance ticket sales for own venues. Real-time velocity (tickets/hour), total sold, remaining capacity ·
Features: tickets_sold_advance, ticket_velocity_per_hour, tickets_vs_median_pct, sellout_probability ·
ML impact: Best real-time predictor. Velocity >2x median = +12% forecast. Feature importance: 0.22 (highest in model).
LIVE
🏅
gov.uk Calendar API
Endpoint: [REDACTED] · Public holidays & school terms API ·
Data: Bank holidays (England, Scotland, NI differ), school term dates by LEA ·
Features: bank_holiday_tomorrow (boolean), school_holiday_active, days_to_next_bank_holiday ·
ML use: Bank holiday eve = +18–22%. School holiday = −4% clubs, +12% family dining. Staggered by region.
LIVE
🎓
University Term Date Scraper
Sources: [REDACTED] · University academic calendar scraper (6 institutions) ·
Features: uni_term_active_{city} (boolean), exam_season_active, freshers_week_active, days_to_term_end ·
ML impact: Durham term end = Klute −62% within 48hrs. Freshers = +42–68%. Feature importance: 0.11 for student venues.
LIVE
💼
SLC Maintenance Loan Dates
Source: [REDACTED] · Student finance payment calendar ·
Features: loan_drop_date (Sep, Jan, Apr), days_since_loan_drop, loan_spending_decay_curve ·
ML use: Spike lasts 10–14 days post-drop. Sep drop = largest (£2,500 avg). Klute +34% in first week after Apr drop.
LIVE
👪
Festival Calendar (Ents24 / Songkick)
Sources: [REDACTED] · Festival & live music calendar APIs ·
Data: Festival dates, locations, capacities, lineup genre tags ·
Features: festival_this_weekend (boolean), festival_capacity, festival_genre_overlap, festival_distance_miles ·
ML use: Glastonbury (210k) = −8–12% every UK city. Parklife MCR (80k) = −22% Factory/Impossible.
PHASE 2
12·e
Transport & infrastructure feeds
🚇
TfL Unified API (Night Tube)
Endpoint: [REDACTED] · Transport disruption API ·
Data: Night Tube status (Victoria, Central, Jubilee, Northern, Piccadilly), service disruptions, planned closures ·
Feature: night_tube_running (boolean). Night Tube suspended = −18% after midnight for London venues.
LIVE
🚌
Nexus Tyne & Wear Metro
Source: [REDACTED] · Metro service status & engineering works ·
Feature: metro_disrupted (boolean), metro_last_service_time. Strike days = −8% Digital, −12% Neptune House.
LIVE
🚂
National Rail / ASLEF / RMT Strike Calendar
Sources: [REDACTED] · Rail network & union strike calendar ·
Feature: rail_strike_active (boolean), tpe_disrupted, lner_disrupted. York Impossible −14% on strike days (relies on Leeds/Harrogate rail).
LIVE
🚗
Council Road Closure Feeds
Sources: [REDACTED] · Roadworks & council event calendar feeds ·
Feature: road_closure_near_venue (boolean). Great North Run = NCL city centre closed but +80% daytime trade.
PHASE 2
🚕
Uber Surge API (Partner)
Endpoint: [REDACTED] · Ride-hailing surge pricing API (partner) ·
Data: Real-time surge multiplier by zone. 2.5x+ surge at 01:00 = −8% late-night trade (people leave early or stay too long). ·
Feature: uber_surge_multiplier, surge_above_2x (boolean). NYE surge reaches 5x.
PHASE 2
12·f
Social & digital signal feeds
📷
Instagram Graph API / Meta Business
Endpoint: [REDACTED] · Social media insights API ·
Data: Post reach, story views, reel plays, saves, shares per venue account. Follower growth. Hashtag volume. ·
Features: ig_reel_reach_7d, ig_saves_announcement, ig_story_poll_yes_pct ·
ML use: Announcement post with >5k saves = +8% baseline. Story poll responses = real-time demand signal.
LIVE
🎬
TikTok Business API
Endpoint: [REDACTED] · Short-form video analytics API ·
Data: Video views, shares, hashtag volume for venue tags, UGC detection ·
Features: tiktok_viral_spike (boolean, >500k views), tiktok_ugc_volume_7d, tiktok_hashtag_growth_pct ·
ML use: Viral moments create 2–4 week demand waves. Feature importance: 0.06 (growing as TikTok adoption increases).
PHASE 2
🔎
Google Search Console + GMB API
Endpoints: [REDACTED] · Search console & business profile APIs ·
Data: “Near me” search impressions, GMB clicks/calls/directions, review velocity and sentiment ·
Features: gmb_searches_7d, gmb_direction_requests, review_score_30d_delta, negative_review_spike (boolean) ·
ML use: Every −0.5☆ drop correlates with −8% walk-in traffic over 60 days. Direction requests = real-time footfall proxy.
LIVE
📰
Meltwater / Brandwatch (Press Monitoring)
Data: Print, online, radio, podcast mentions per venue. Sentiment analysis. Reach estimates. ·
Features: press_mentions_7d, press_sentiment_score, press_reach_estimated ·
Cost: £8–15k/yr · Press mentions have 7–14 day decay curve on footfall. TimeOut review = ~8k reads, Manchester Evening News feature = ~180k reach.
PHASE 2
💬
WhatsApp Business API (Promoter Lists)
Endpoint: [REDACTED] · Messaging platform API ·
Data: Broadcast open rates, reply rates, booking enquiry volume per venue ·
Features: whatsapp_open_rate, whatsapp_enquiry_count, whatsapp_enquiry_vs_avg ·
ML use: Open rate >60% = high demand night. Enquiry-to-booking conversion: 72%. 14,200 subscribers across group.
LIVE
12·g
Internal EPOS & ops feeds
💳
Zonal EPOS (Revenue Data)
Source: Zonal iServe EPOS system · 412 terminals across 68 venues ·
Data: Transaction-level: item, time, terminal, operator, void, discount, payment method, covers ·
Update: Real-time (WebSocket push) + nightly batch reconciliation ·
ML role: THE training target. Every external factor is correlated against daily venue revenue from EPOS. 3+ years of history = 68 venues × 1,095 days = 74,460 training rows.
PENDING
💵
Sage Line 50 (Financials)
Source: Sage API bridge · GP%, COGS, overhead, rent, rates per venue ·
ML role: Margin modelling. Revenue alone is misleading — need GP% to calculate EBITDA impact of marketing uplift. A £10k uplift at 71% GP = £7.1k profit. At 64% GP = £6.4k. Sage provides the multiplier.
PENDING
👥
RotaCloud / Fourth (Labour)
Source: RotaCloud API + Fourth Analytics · Rota data, actual hours, labour cost per hour, staff count per shift ·
ML role: Labour as % of revenue is the intervention target. Model predicts quiet nights → recommends fewer staff. Saves £340k/yr across group. Feature: labour_pct_actual, staff_count_shift, labour_vs_target_delta.
PENDING
12·h
Integration architecture
Data Pipeline ArchitectureHow feeds become predictions
1
Collection Layer
Runtime: Node.js workers on Render (cron + webhook triggers)
Schedule: Hourly: weather, EPOS, transport status. Daily: economic, flights, social. Weekly: Eurostat, GfK, ONS. Quarterly: Ibestat, DTCM.
Storage: PostgreSQL (Neon) for structured time-series. Redis for real-time cache. S3 for raw API response archive.
Error handling: Circuit breaker per feed. If Met Office is down, failover to Open-Meteo within 30s. Stale data alert if any feed >2x expected refresh interval.
2
Feature Engineering
Runtime: Python (pandas + scikit-learn) on scheduled worker
Process: Raw API data → 70+ engineered features per venue per day. Includes: rolling averages, YoY comparisons, boolean flags, interaction terms (e.g., payday × dry weather = super feature).
Output: Feature matrix: 68 venues × 15 days forward × 70+ features = 71,400 feature vectors per batch
Validation: Feature distribution checks. Outlier detection. Missing value imputation (forward-fill for economic, interpolation for weather).
3
ML Prediction
Model: Proprietary ML ensemble. Trained on 3+ years of EPOS × external factors.
Retraining: Weekly (Sunday overnight). Uses last 13 weeks of data with exponential decay weighting.
Output: Baseline revenue prediction per venue for next 15 days. Confidence interval (±2.4% at 97.6% confidence).
Performance: 96.9% accuracy over last 13 weeks. 68 venues × 15 days predicted in <2 seconds.
Feature importance: Proprietary weighting across 70+ features. Top factors include demand signals, entertainment, temporal patterns, weather, and economic indicators. Exact weights are confidential.
4
LLM Reasoning + Agent Execution
LLM: Multi-provider LLM (6 provider failover). Interprets ML outputs, explains WHY, generates natural language briefings.
Agents: 25+ autonomous agents. Budget reallocation, campaign creation, GM notifications, staffing adjustments, content generation.
Closed loop: Agent actions → results feed back into EPOS → ML model retrains → better predictions → better agent actions.
Human override: Any agent action >£5k or “pause venue” requires GM approval via WhatsApp confirmation.
12·i
Cost summary
| Feed Category | Count | Free | Paid | Annual Cost | Status |
|---|---|---|---|---|---|
| Weather & Environment | 3 | 3 | 0 | £0 | LIVE |
| Aviation & Tourism | 8 | 4 | 4 | £57–138k | MIXED |
| Economic & Macro | 7 | 7 | 0 | £0 | LIVE |
| Events & Entertainment | 8 | 5 | 3 | £0–2k | MIXED |
| Transport & Infrastructure | 5 | 3 | 2 | £0–5k | MIXED |
| Social & Digital | 5 | 3 | 2 | £8–15k | MIXED |
| Internal EPOS & Ops | 3 | 3 | 0 | £0 | PENDING |
| Total | 39 | 28 | 11 | £65–160k | 72% FREE |
Phase 1 (MVP): 22 free feeds + EPOS data = 80% of model accuracy for £0 marginal cost.
Phase 2: OAG + ForwardKeys + Meltwater add the remaining 20% precision for £65–160k/yr. ROI: model accuracy improvement from 93% → 97% = estimated £400k incremental revenue attribution accuracy across group.
Phase 2: OAG + ForwardKeys + Meltwater add the remaining 20% precision for £65–160k/yr. ROI: model accuracy improvement from 93% → 97% = estimated £400k incremental revenue attribution accuracy across group.