Flow

Managing order flows in real time

Centralise your multi-platform orders, manage activity peaks and avoid late penalties with strategies proven by the best restaurants.

The challenge

The chaos of multi-platform orders

Uber Eats, Deliveroo, Just Eat, phone orders, dine-in customers. When four order sources simultaneously feed your kitchen, the slightest hiccup can derail the entire chain. Here are the problems faced by 80% of multi-platform restaurants.

Three tablets, three ringtones

Each platform has its own tablet with its own notifications. When all three ring at the same time during a rush, stress rises and errors multiply. One missed validation on a tablet and the order is automatically cancelled after the timeout.

No overview

Impossible to know at a glance how many orders are in progress, which ones are late and how many are arriving in the next 15 minutes. Without visibility, you navigate blind and react instead of anticipate.

Cascading errors

An order prepared with an error generates a complaint. The time spent handling the complaint slows down subsequent orders. These delays generate new complaints. The vicious cycle sets in within minutes.

Team fatigue

Managing multiple flows simultaneously is cognitively exhausting. After 3 hours of service with relentless orders, concentration drops and the error rate increases by 40 to 60%. Fatigue is the primary factor behind delivery errors.

Centralisation

Centralising all your flows in one place

The solution to multi-platform chaos is centralisation. Instead of juggling between multiple tablets and interfaces, consolidate everything in a single system that gives you a clear and actionable overview. Here are the pillars of successful centralisation.

Single dashboard

Group all orders from all platforms in a single interface. No more back and forth between tablets. Every order is visible with its status, deadline and origin platform. One screen to monitor instead of three or four.

Tablet management

Assign tablets by station: one for the kitchen, one for packaging, one for delivery coordination. Each station only sees the orders that concern it, in priority order. Less noise, more efficiency.

Smart print routing

Order tickets print automatically on the right printer: starters to the cold station, hot dishes to the cooking station, drinks to the bar. No more manually sorting tickets — each station receives exactly what it needs to prepare.

Alerts and priorities

The system identifies orders close to their deadline and escalates them. A specific audio alert triggers when an order risks being late. You intervene before the problem occurs, not after.

Refusal management

Refusing intelligently without destroying your ranking

Refusing an order is not a failure — it is sometimes the best decision to protect the quality of your other orders. But every refusal has an algorithmic cost. The challenge is knowing when to accept and when to refuse, and how to minimise the impact on your visibility.

When refusing is the right decision

  • Your kitchen is already at capacity and accepting would degrade the quality of all current orders
  • The ordered item is out of stock and you cannot offer an acceptable alternative
  • The estimated delay far exceeds the maximum acceptable time for the customer
  • The order is clearly fraudulent or nonsensical

The impact of refusals on your ranking

  • An acceptance rate below 85% triggers a degradation of your visibility on the platform
  • One-off refusals during peak hours are better tolerated than regular refusals
  • Uber Eats measures your rate over 28 rolling days — a bad day dilutes quickly
  • Deliveroo takes the reason for refusal into account: 'too busy' is better perceived than 'item unavailable'
Peak hours

Surviving rush hours without sacrificing your score

The 12pm–1:30pm and 7pm–9pm slots often represent 70% of your delivery revenue. They are also the moments when errors are most costly. Good preparation and a few strategic adjustments make the difference between a controlled service and a catastrophic one.

Adjust preparation times

During peak hours, increase your displayed preparation times by 5 to 10 minutes on each platform. This reduces incoming flow and gives you a buffer to deliver on time. Better to promise 40 minutes and deliver in 35 than to promise 25 minutes and deliver in 35.

Simplify the menu during peak hours

Deactivate complex items (more than 10 minutes preparation) during the 12pm–1:30pm and 7pm–9pm slots. Keep only the quickly prepared dishes with high satisfaction rates. You'll serve more customers with less stress.

Stagger platform openings

Don't open all platforms at the same time. Start with Uber Eats at 11:30am, add Deliveroo at 11:45am. This spreads the initial flow and avoids the brutal spike that overwhelms the kitchen in the first minutes of service.

Plan staffing

Analyse your order data from the past 4 weeks to identify peak time slots. Schedule an extra cook for critical hours. The cost of this backup is more than offset by the reduction in cancellations and penalties.

Late penalties

How platforms penalise delays — and how to avoid them

Preparation delays are the number one factor in ranking degradation on delivery platforms. Understanding how each platform measures and penalises delays is essential to protect your visibility.

Uber Eats

Uber Eats measures your actual preparation time (from the moment you accept to the moment the dish is ready). If you regularly exceed the announced time, your ranking drops progressively. Beyond a 15-minute average delay, your restaurant may be temporarily deactivated during peak hours.

Deliveroo

Deliveroo uses a punctuality score system. Every order delivered late lowers this score. Below a certain threshold, your restaurant no longer appears in default search results — the customer must explicitly search for you by name to find you.

The time-buffer strategy

The most effective technique to avoid penalties is to add a buffer to your estimates. If your actual preparation time is 15 minutes, display 20–25 minutes on the platform. This 5 to 10-minute buffer lets you absorb unexpected situations (complex order, ingredient to fetch, unexpected peak) without ever exceeding the announced deadline.

Bonus: customers who receive their order ahead of the announced deadline are significantly more satisfied and leave better reviews. An announced deadline of 30 minutes with delivery in 22 minutes generates more satisfaction than an announced deadline of 20 minutes with delivery in 22 minutes — even though the actual time is identical.

Automation

Automate to improve reliability

Every manual task is a potential source of error. Automation doesn't replace humans — it frees them from repetitive tasks so they can focus on what really matters: kitchen quality and customer satisfaction.

Automatic confirmation

Configure automatic order acceptance during off-peak hours (2pm–6pm). This improves your average response time and acceptance rate without extra effort. Keep manual validation only during peak hours to control the flow.

Kitchen display screen (KDS)

Replace paper tickets with a digital kitchen display. Orders appear in real time, colour-coded by urgency. The cook validates each step with a gesture and the system automatically updates the status across all platforms.

Delivery coordination

When a dish is marked as ready in the kitchen, the system automatically notifies the platform to trigger the courier's arrival. Optimised timing reduces the courier's waiting time (and therefore total delivery time) by 5 to 8 minutes on average.

Automated daily reports

Receive a summary of your day every evening: number of orders, average preparation time, cancellation rate, late orders. This data lets you adjust your organisation day after day without having to compile it manually.

FAQ

Frequently asked questions

How many orders per hour can a restaurant handle for delivery?
This depends on your kitchen capacity, but on average, a restaurant with 2 cooks can handle 15 to 25 orders per hour for delivery. Beyond that, quality and deadlines deteriorate. If you regularly hit your limit, it is better to increase preparation times rather than refuse orders.
Do you need specific equipment to centralise orders?
At a minimum, you need a tablet or computer with multi-platform management software. Solutions range from simple spreadsheets (free but manual) to dedicated tools like Deliverect, Orderbird or integrated POS systems. The investment ranges from £0 to £200 per month depending on the solution chosen.
How do you avoid late-delivery penalties without refusing orders?
The most effective strategy is to increase your displayed preparation times during peak hours. Move from 20 minutes to 30–35 minutes. The customer sees a longer delay but receives their order on time. Your punctuality score improves and your ranking rises. Paradoxically, displaying a longer time can increase your orders if you consistently deliver ahead of schedule.
What is the best time to adjust your delivery settings?
Make your adjustments on Sunday evening or Monday morning, before the start of the week. Analyse the previous week's data, identify problematic time slots, and adjust preparation times, active menus and opening hours accordingly. Avoid modifying your settings during service — it creates confusion.

Join the FluxResto community on WhatsApp

Free · Exclusive content · 0 spam · Cancel anytime

Join the channel

Instant access · WhatsApp