Read the cookie policy

Quick service · Restaurants · Bakeries

From stock to service, one system.

Enskel forecasts every service, prepares purchasing and staff schedules, and tracks margins, from one site to a whole group.

30 minutes by video call with Steven or Baptiste, Enskel’s co-founders.

In this demo, Julien runs three restaurants. Today’s alerts, Saturday’s forecast, the rota, the order, the margin per dish and the month’s result.In French, with French subtitles · sample data
Video transcript and description
  1. On screenOn a white background, a large clock goes from 14:19 to 14:20, then the group dashboard appears (‘Group · 14:20’): 364 customers served, 8.9% fewer than forecast; €5,163 revenue excl. VAT (2/3 sites); average spend €14.18 excl. VAT; 3 sites needing attention, 14 open alerts, 3 of them urgent. The site table (customers served / forecast) shows Vincennes: incomplete data, Montreuil: 132 / 145, Nation: 232 / 254.

    Voice-over‘Friday, 14:20. Julien runs three restaurants. At a glance, he sees how each one is doing:’

  2. On screenThe ‘Variance’ column fills in (Montreuil −9.2%, Nation −8.8%) and the ‘Needs attention’ panel lists three alerts: ‘Vincennes · data — POS feed interrupted’, ‘Nation · food safety — Non-compliance: cold room CF2 at 7.8 °C’ and ‘Montreuil · activity — Activity below forecast today at 14:15 (−9.2%)’.

    Voice-over‘a silent till at Vincennes, a cold room running too warm at Nation.’

  3. On screenThe group’s ‘Today’ view (tabs Yesterday, Today, Tomorrow, Week, Month; ‘Group · 3 sites · Fri 2 Oct · 14:20’) with the summary ‘At 14:20, the group has had 364 customers, 8.9% fewer than forecast. 3 sites have an issue to resolve.’ The background turns black and the ‘Tomorrow’ tab (Saturday 3 October) opens; under ‘Coming up’: ‘Event · Sat 3 Oct · Nation — Match screened on Place de la Nation (evening)’.

    Voice-over‘And what awaits him tomorrow evening: a match on Place de la Nation.’

  4. On screenNation’s forecast: a chart of customers per day, up to 16 October, where Saturday’s bar stands out in red, then the figure for Saturday 3 October at Nation: ‘1,094 customers forecast’, ‘between 1,015 and 1,210’.

    Voice-over‘Enskel has already factored it in: 1,094 customers expected on Saturday,’

  5. On screenThe ‘Why this figure’ table breaks down the forecast, in customers: learned average level 677, Saturday effect +165, recent level +32, match screened on Place de la Nation (evening) +140 (underlined row), rounding −1, automatic calculation 1,013, manual adjustment ‘delivery platform promotion (free delivery)’ +81, final forecast 1,094.

    Voice-over‘140 of them thanks to the match.’

  6. On screenOn a black background, Saturday unfolds in bars, one per half hour, then zooms in on the evening: 18:00 35 customers, 18:30 48, 19:00 66, 19:30 82, 20:00 80, 20:30 66, 21:00 47, 21:30 40, 22:00 25, 22:30 28. The 20:00 bar stands alone: ‘80 customers, cautious need, 20:00–20:30’.

    Voice-over‘The day takes shape, slot by slot. Between 20:00 and 20:30, expect 80 customers.’

  7. On screenSlot card ‘20:00–20:30 · evening’ (Saturday 3 October · 80 customers forecast, cautious need), tagged ‘Need 15 people’. People per station: fryer, 124 min of work, minimum 1, need 5, scheduled 4.0; assembly, 62 min, minimum 1, need 3, scheduled 2.0; counter, 106 min, minimum 1, need 5, scheduled 4.0; prep, 40 min, minimum ‘—’, need 2, scheduled 1.0. Alongside, a ‘cautious need’ column shows 15, then an ‘on the rota’ column shows 11, with a shortfall of 4 in red; the tags ‘Scheduled 11’ (in red) and ‘Manager required’ appear.

    Voice-over‘To be safe, it will take 15 people. The rota only has 11.’

  8. On screenChicken fillet, ‘Quantity calculation’ for 3 Oct (Volailles du Val), in kg: forecast requirement 251.3, plus safety margin 25.1, minus stock available when the requirement starts 0.0, plus forecast usage before the requirement, beyond stock 89.8, minus deliveries confirmed in time 250.0: 116.2 left to order before rounding. Order already placed: CMD-1001-VDV, 250.0 kg, confirmed, available Sat 3 Oct at 08:30, before the requirement. The total grows larger: ‘116.2 kg short for Saturday’. Back on a white background: ‘The shortfall shows up on Friday.’, next to the same card dated ‘Fri 2 Oct · 14:20’.

    Voice-over‘In the kitchen, Saturday is 116 kg of chicken short. Julien is alerted, and he decides.’

  9. On screenChicken fillet stock: counted at 99.0 kg at 10:30 (‘Count validated’), it goes down with each sale according to the recipe portions (chicken fillet per sale, gross: 160 g for the Burger Crousti, 200 g for the Tenders x5, 120 g for the Tenders x3) to 40.3 kg ‘projected’ at 14:20, i.e. ‘41% of count’. The log adds the line ‘Sales (recipes) · theoretical, up to 14:20’: −58.7.

    Voice-over‘Stock updates in real time: every sale is deducted, recipe by recipe.’

  10. On screen‘Week closed · chicken fillet’, 21 to 27 Sept, in kg: opening stock 120.0; goods received +640.0; theoretical sales (recipes) −672.0; recorded waste −2.0; theoretical stock before count 86.0. Two bars compare theoretical (86.0) and counted (84.0): ‘−2.0 kg’, ‘variance to explain’. The ‘−2.0 kg to explain’ card (tagged ‘Variance’) lists leads marked ‘Hypothesis’: conversions and yields, sales and POS reconciliation, deliveries and documents, waste and authorised removals, and, highlighted, recipes and versions: ‘Burger Crousti: recipe card changed on Mon 21 Sept’.

    Voice-over‘At stocktake, theoretical against counted: 86 kg expected, 84 counted. And every variance is tracked, week after week.’

  11. On screenAfter a brief ‘Burger Crousti’ title, the price of chicken fillet (contract price · Volailles du Val): €6.40 per kg since 1 March 2026, €6.70 since 15 Sept 2026, i.e. +4.7%. A table of recipe cards scrolls past, blurred, during the recalculation.

    Voice-over‘When chicken goes up 4.7%, every recipe card recalculates:’

  12. On screen‘Burger Crousti’ card (tabs Eat in, Takeaway, Delivery, Click and collect; Eat in selected): selling price €7.18 excl. VAT (€7.90 incl. VAT); ingredient cost goes from €1.77 to €1.82 and the final margin from €1.94 to €1.89. ‘Where the €7.18 excl. VAT from a sale goes’: ingredients 25.6%, labour 14.0%, selling costs 7.7%, overheads 26.4%, final margin 26.3%.

    Voice-over‘Julien sees where every euro of a burger goes.’

  13. On screenOn a black background, ‘Checks · today’ (Nation · Fri 2 Oct): cold room CF1 at 3 °C, cold room CF2 at 7.8 °C, both read at 08:00; CF2 switches to ‘Out of range’ (range: 0 to 4 °C). Its history for the day: 08:00, 7.8 °C, source: sensor; 14:00, scheduled, not done; 16:00, upcoming. Under ‘Batches’, the 30 Sept coleslaw (batch L-0930-COL) switches to ‘In quarantine’, and a task appears: ‘Make a new 6 kg batch of coleslaw for tonight and tomorrow’ (To redo · 1) — ‘The 30 Sept coleslaw is in quarantine: it does not count as available. The new batch keeps until Mon 5 Oct, with no extension possible.’

    Voice-over‘And this morning’s cold room? The check is logged, the coleslaw blocked, a batch to remake.’

  14. On screenOn a black background, ‘The margin?’. Then, for the Bucket 9 tenders at Nation, service on Thu 1 Oct, the planned margin at menu price (€700.00) melts away line after line: discounts and refunds −70.00; purchase price effect −20.00; batch difference −15.00; remakes −6.70; end-of-service waste −26.80; confirmed commission −93.00, in red; unexplained variance 0.00, ‘Bridge closed to the cent’. Margin after costs: €468.50.

    Voice-over‘And the margin? Every point lost has its reason: discounts, purchase prices, portions thrown away, commission.’

  15. On screenOn a white background, the same lines form the ‘From planned to actual’ table (Bucket 9 tenders · Croustille Paris Nation · service on 1 Oct), dated ‘Fri 2 Oct’; the date ‘Sat 31 Oct’ is crossed out.

    Voice-over‘Julien sees them the next day, not at the end of the month.’

  16. On screen‘Profitability’ (Croustille Paris Nation · September): ingredients account for 23.3% of sales, against 21.8% planned by the recipe cards, i.e. +1.4 pts. Two bars compare the recipe-card plan (€63,355) and the actual (€67,550); the variance appears: €4,195, +1.4 points.

    Voice-over‘In September, ingredients cost 1.4 points more than the recipe cards planned: €4,195.’

  17. On screenA red bar marked ‘1%’ and the amount ‘€2,901’: ‘one margin point · September · Nation’.

    Voice-over‘And at Nation, one point of margin is worth more than €2,900.’

  18. On screenFour panels scroll past. ‘scheduling’: the Saturday evening slot. ‘portion tracking’: where the Bucket 9 tenders portions went, service on 1 Oct — 100 served to customers, 2 set aside then r… (label cut off on screen), 8 thrown away at end of service (in red), 0 still in stock, 0 unaccounted for. ‘menu profitability’: ‘Bucket 9 tenders weighs most on margin.’ ‘margin by channel’: margin after costs in September, 69.7% eat in, 69.7% takeaway, 61.7% click and collect, 42.6% delivery (flagged in red).

    Voice-over‘So he acts in the right place: the team, production, the menu, the channels.’

  19. On screenWhite screen, then three indicators follow one another: Saturday’s forecast, 1,094 customers · Nation; staffing need, 15 people at 20:00, Saturday; suggested order, 15 items · Nation.

    Voice-over‘And every morning, the longest part is already done: the forecast, the staffing need, the order.’

  20. On screen‘Daily tasks’ list: twice ‘Enter the check: cold room CF1’ (Enter button), twice ‘Approve a suggested order’ (Approve button), ‘Publish the week 41 rota’ (Publish button), ‘+1 task · choose a site’, and the note ‘Croustille Vincennes: POS only, no on-site tasks until setup is done.’ The cursor clicks the first ‘Approve’ button.

    Voice-over‘Julien checks it over and approves.’

  21. On screenThe chain rebuilds from left to right: customers forecast per half hour (‘80 from 20:00 to 20:30’), the need for 15 people against the 11 scheduled, the 116.2 kg chicken fillet order, the 160 g of fillet per Burger Crousti, the margin for Thu 1 Oct (€468.50), then the month’s result.

    Voice-over‘Everything is connected: the forecast drives the rota and the order, the recipe drives the margin, the margin drives the result.’

  22. On screenThe full app, with its menu (Today, Forecasting, Stock, Recipes, HACCP, Scheduling, Profitability, Sites), moves from one module to the next: Forecasting (customers per half hour, 82 at 19:30 in red), Scheduling (20:00–20:30 slot), Stock (chicken fillet: 40.3 kg projected stock, 40.3 kg usable), Recipes (Burger Crousti), HACCP (CF1 at 3 °C ‘Compliant’, CF2 at 7.8 °C) and Profitability (23.3%). Then the title ‘From stock to service, one system.’ and the ‘Monthly result’ page (Results › Croustille Paris Nation; tabs September and October · in progress): €303,830 revenue excl. VAT, ingredients 23.3% of sales, and a third indicator at 30.5% whose label runs off the frame.

    Voice-over‘Forecasting, scheduling, stock, recipes, hygiene, margins: it all fits in one tool, right through to the month’s result.’

  23. On screenThree access levels follow one another. ‘Julien · group’: the dashboard for the three sites. ‘Manager · restaurant’: ‘Today’, Croustille Paris Nation, 14:20 — ‘At 14:20, your restaurant has had 232 customers, 8.8% fewer than forecast. Your restaurant has one issue to resolve.’; €3,380 revenue excl. VAT, average spend €14.57 excl. VAT; needs attention: ‘Nation · food safety — Non-compliance: cold room CF2 at 7.8 °C’. ‘Team member · shift’: a smartphone held in the hand shows ‘Today · Nation · Fri 2 Oct — You are on the fryer station until 20:00. 3 checks to do during your shift.’, the 12:00–20:00 shift · fryer (20 min break), and the checks to do: cold room CF2 at 14:00 ‘Overdue’, CF1 at 16:00 ‘Upcoming’, CF2 at 16:00 ‘Upcoming’; at the bottom, the tabs Today, Rota, HACCP, Stock.

    Voice-over‘And everyone has their own access: Julien, his whole group; the manager, their restaurant; and each team member, on their phone, their rota and their checks.’

  24. On screenEnd cards on a white background: ‘Enskel recommends.’ then ‘You decide.’

  25. On screenThe Enskel symbol appears alone in the centre (the black diamond with its white E and red tip), then the tagline ‘See what’s coming. Decide sooner.’, a ‘Book a demo’ button and the address enskel.com, until the end of the video (2:34).

01 / THE DAY

A day, seen in advance.

From the 06:00 bake to evening service.

  1. The bake

    Quantities of every item, forecast the day before.

    140 croissants, 60 baguettes.

  2. Purchasing

    What lunch and dinner need, already calculated.

    Cream: 2 L to order before 15:00.

  3. The rush

    The right team in place, station by station.

    6 team members on shift.

  4. The checks

    Temperatures and cleaning, logged on the tablet.

    Cold room: 3.1 °C, compliant.

  5. Evening service

    The brigade in place, matched to the dinner forecast.

    84 covers tonight.

sample data

02 / SEE IT COMING

Sales, weather, events.

The forecast takes into account what changes a service, and tells you how far off it could be.

Tuesday · 10:00

What could change service

Bastille

WeatherRain from 12:00: terrace closed, 12 fewer covers.

Lunchtime weather

Forecasting · Bastille

  1. : 17 : 17
  2. : 27 : 25
  3. : 40 : 36
  4. : 52 : 48
  5. : 33 : 31
  6. : 21 : 21
190178 covers forecast for lunch within 7% · sample data

See the forecast

03 / THE DIAGNOSIS

Too many tools, no shared view.

Each tool keeps its own data and costs its own subscription. A good forecast, on the other hand, can be measured.

What stacked-up tools cost

37%

of chains cite scattered data and tools as a barrier to customer experience.

168 quick service chains with more than 20 sites each, over 94,000 sites in total, in 2026.

SourceQu, 2026 Restaurant Technology Benchmarkqubeyond.com

€15,800

a year: the average budget of an independent restaurateur for digital equipment.

Nearly 900 restaurateurs and hoteliers surveyed in France, at the Food Hotel Tech trade show, on 19 and 20 March 2025.

SourceFood Hotel Tech 2025 surveytendancehotellerie.fr

What a good forecast changes

  • up to6%

    lower food cost

  • 5–15%

    lower staff costs

  • up to50%

    less food waste, the project’s stated target

These results come from another forecasting tool, not Enskel.

Measured with the Prognolite tool in 12 restaurants from six Swiss and German groups, in 2019.

SourceEuropean Commission, Prognolite projectcordis.europa.eu

04 / PREPARE EVERYTHING

Six modules, one system.

The forecast sets out the service; stock, the rota and prep follow.

01

Forecasting

Sales for every service, seen in advance.

Saturday: 562 covers, within 7%.

sample data

  1. 02

    Stock

    Purchasing, calculated from the forecast.

    Cream: 2 L to order before 15:00.

  2. 03

    Recipe costing

    Every dish to the gram, with its food cost.

    Burger: €2.84 food cost.

  3. 04

    Scheduling

    Shifts, matched to forecast footfall.

    Saturday: two more people.

  4. 05

    HACCP

    Temperatures, cleaning and traceability.

    Cold room: 3.1 °C, check done.

  5. 06

    Profitability

    Theoretical food cost against actual.

    Food cost: 28.6% of sales.

See the six modulesThree packages, three prices per site: see pricing

05 / FROM HEAD OFFICE TO EVERY SITE

Each site sees its day. Head office sees them all.

Each site prepares its service on its own tablet. Head office sees the whole group: forecasts, purchasing, overdue checks, margins, and the detail of every site.

  • Bastille

    190 covers forecast for lunch

    1. 11:30 : 17
    2. 12:00 : 27
    3. 12:30 : 40
    4. 13:00 : 52
    5. 13:30 : 33
    6. 14:00 : 21
  • Batignolles

    142 covers forecast for lunch

    1. 11:30 : 9
    2. 12:00 : 24
    3. 12:30 : 37
    4. 13:00 : 40
    5. 13:30 : 22
    6. 14:00 : 10
  • Confluence

    96 covers forecast for lunch

    1. 11:30 : 6
    2. 12:00 : 16
    3. 12:30 : 24
    4. 13:00 : 27
    5. 13:30 : 15
    6. 14:00 : 8
  • Bastille 190
  • Batignolles 142
  • Confluence 96

HACCP Confluence: 2 HACCP checks overdue.

Covers forecast per half hour · three sites · Tuesday 27 October sample data

06 / SECTORS

One system, three trades. Each with its own words and hours.

  • Quick service

    Saturday’s rush, forecast from Monday.

    Châtelet · Saturday 31 October

    1. 11:30 : 9
    2. 11:45 : 14
    3. 12:00 : 22
    4. 12:15 : 31
    5. 12:30 : 41
    6. 12:45 : 38
    7. 13:00 : 33
    8. 13:15 : 24
    9. 13:30 : 15
    10. 13:45 : 8
  • Restaurants

    Evening prep, matched to sales and the weather.

    • Weather Saturday: 17 °C, sunny, terrace open.
    • Events Saturday: match at 21:00.
  • Bakeries

    Each bake, baguette by baguette.

    140 croissants, 60 baguettes.

sample data

07 / SEE FOR YOURSELF

The product, as it is.

Its screens, its method, its founders.

  • 01

    The product

    The tablet at the pass, the manager’s phone, the head office dashboard.

    See the product

    Lunch service

    Bastille · 10:30

    1. 11:30 : 17
    2. 12:00 : 27
    3. 12:30 : 40
    4. 13:00 : 52
    5. 13:30 : 33
    6. 14:00 : 21
    Forecast covers per half hour 190 in total · peak at 13:00 · within 7% sample data
  • 02

    The method

    Forecasting sales in six steps, with a worked example.

    Read the method

    Forecasting restaurant sales

    1. Start from the right unit
    2. Set an honest baseline
    3. Adjust for the calendar
    4. Add the weather and the terrace
    5. Measure the error, and compare it with the baseline
    6. Turn it into orders and the rota
  • 03

    The team

    Steven Poivre and Baptiste Fondanèche, Enskel’s co-founders.

    Meet the team

    Portrait of Steven Poivre
    Portrait of Baptiste Fondanèche