How to automate your restaurant with n8n: bookings, orders, inventory, and collections (2026)

Automating a restaurant with n8n can save 20-30 hours of weekly admin work and integrates Fudo, MercadoPago, and Tiendanube into a single flow. Whether it pays off is a calculation with your own numbers: how many tickets a day you handle and how many hours go into manual tasks today.

TL;DR

• n8n is the most used self-hosted automation platform in AR restaurants 2026 (fixed cost vs Zapier which charges per task).
• The 5 highest-ROI processes: POS-delivery inventory sync, WhatsApp bookings, automatic orders, post-visit reviews, daily reporting.
• Typical stack: Fudo or Maxirest (POS) + PedidosYa/Rappi + MercadoPago + Tiendanube + WhatsApp + n8n.
• Hypothetical scenario: 80-seat bodegón with delivery, to see how the stack comes together and what is worth measuring.
• Cost: no fixed plans. It is set in a proposal with scope, phases, and timelines; third-party infrastructure and licenses are separate.

Why n8n is the best option for Argentine restaurants in 2026

n8n is an open-source automation platform (Sustainable Use license) that lets you connect APIs and build flows with a visual editor. In AR restaurants it gained ground over Zapier for three concrete reasons:

1. Fixed cost. n8n Community runs on a Hostinger VPS for USD 10-15/month with no task limit. Zapier, at 5000 tasks/month, costs USD 73/month.
2. Self-hosted. Your customer data (emails, phone numbers, tickets) stays on a server you control. Important for complying with Law 25.326.
3. Flexibility for the regional stack. You can call APIs for Fudo, Tusfacturas, MercadoPago, and AFIP via HTTP Request without waiting for the provider to build the connector.

The 5 most profitable processes to automate in restaurants

Not everything is worth it. This is the list prioritized by the impact these processes usually have in a restaurant with active delivery.

1. Inventory sync between POS and delivery apps

The pain: the waiter sells the last potato pie in the dining room, but PedidosYa keeps showing it as available. An order comes in, you cancel it, you lose the customer and eat the commission.

The automation: each sale in Fudo triggers a webhook → n8n updates inventory on PedidosYa Partner Portal, Rappi, MercadoLibre Delivery, and Tiendanube. If the product hits 0, it pauses across all channels automatically.

Savings: 5-10 cancellations avoided/week, +3-5% in delivery NPS.

2. WhatsApp bookings and reminders

Booking no-shows cost 8-15% of weekend revenue. n8n connects Google Calendar, a web form, or TheFork and sends via WhatsApp (via Evolution API or Cloud API):

1. Confirmation when booking.
2. Reminder 24 hours before.
3. Link to cancel if they can't make it.
4. Review request 2 hours after the meal.

Illustrative example, not a guaranteed result: bringing no-shows down from 12% to 4%. Measure it with your own bookings.

3. Automatic orders from WhatsApp to POS

Customer sends a message to the restaurant's WhatsApp → AI bot (GPT-5 or Claude) understands the order, validates inventory in Fudo, calculates total, sends a MercadoPago payment link, and when payment is confirmed it creates the ticket in Fudo and sends the order to the kitchen via fiscal printer or Telegram to the chef.

It's the most profitable flow if your location handles 30+ owned delivery orders per day.

4. Automated reviews and online reputation

Post-meal, n8n sends a WhatsApp with a link to Google Reviews (if they internally rated >=4) or to a private complaint form (if they rated <=3). It filters out negatives before they reach Google.

Typical increase in Google Reviews: from 2-3 per month to 25-40 per month with average rating 4.6+.

5. Daily reporting and register closing

Each night at 23:59, n8n queries Fudo, MercadoPago, Tiendanube, and delivery apps, builds a PDF with sales, average ticket, top products, commissions paid, and sends it via WhatsApp to the owner. The owner sees the day without stepping into the restaurant.

Savings: 1 hour per day for the manager + full visibility for the owner.

n8n workflow step by step: Fudo + PedidosYa + WhatsApp sync

Full flow you can replicate. Assumes self-hosted n8n and Evolution API running.

1. Trigger: Webhook node listening for a POST from Fudo when a ticket is closed.
2. Set node: extract ticket_id, products, total, customer_phone.
3. HTTP Request node to the Fudo API to fetch updated stock for the sold product.
4. IF node: if stock <= 2 units, branch to pause the product.
5. HTTP Request node to the PedidosYa Partner API with PATCH method to mark the product as unavailable.
6. HTTP Request node to Tiendanube to update inventory there too.
7. HTTP Request node to Evolution API (/message/sendText/{instance}) to alert the manager by WhatsApp: "Milanesa napolitana sold out, paused in delivery".
8. PostgreSQL node: save the event in the stock_events table for reporting.
9. Respond to Webhook node: 200 OK to Fudo.

Build time depends on how each API behaves (Fudo, PedidosYa, Tiendanube). In a project we set it in the proposal, separating our work from what depends on each provider.

Recommended stack for Argentine restaurants in 2026

The 5 tool categories you need well integrated.

POS (point of sale)

1. Fudo — leader in AR, mature REST API, integrates with delivery natively.
2. Maxirest — strong in professional restaurants, decent API.
3. Bistrosoft — older but still in use, limited API.

Delivery

1. PedidosYa Partner API — has order webhooks and menu updates.
2. Rappi Partners — similar API, slower.
3. MercadoLibre Delivery — growing in AR 2026.

Payments

1. MercadoPago Checkout Pro — confirmed-payment webhook, 15-minute integration.
2. Modo — for in-person QR payments.
3. Uala Bis — growing alternative.

Communication

1. WhatsApp via Evolution API or Cloud API — main channel.
2. Chatwoot — dashboard for human agents when the bot doesn't know.
3. Email Gmail API — for confirmations and invoices.

Analytics and reporting

1. Looker Studio — free, connects to PostgreSQL via connector.
2. PostgreSQL on Hostinger — central database where n8n stores everything.

Hypothetical scenario: 80-seat bodegón, LATAM

Suppose a traditional bodegón in Almagro, 80 seats, 120 delivery orders/day across PedidosYa and Rappi, monthly revenue USD 45,000. This is not a documented case: it is an example to show how the stack comes together and which numbers are worth measuring.

Pains before implementation:

• Manager closed the register for 1.5 hours every night, adding up Excel sheets.
• Outdated delivery inventory: 8-12 cancellations/day.
• Weekend booking no-shows: 15%.
• Google Reviews stuck at 4.1 with 38 total reviews.

What gets implemented:

1. Deploy n8n + Evolution API + PostgreSQL on Hostinger KVM 4 (USD 15/month).
2. Connect Fudo API + PedidosYa + Rappi + MercadoPago.
3. 6 production workflows: inventory sync, bookings/reminders, nightly reporting, review requests, low-stock alerts, pending-order collections.

What could change in 60 days, under these assumptions:

• Delivery cancellations: from 10/day to 2/day.
• No-shows: from 15% to 5%.
• Google Reviews: from 38 to 112, rating 4.6.
• Manager register-closing time: from 90 min to 10 min.
• Average delivery ticket: +12% (bot suggests sides).

These are sizing assumptions, not guaranteed results: the return is calculated with your restaurant's own numbers before you decide.

How much it costs to automate your restaurant with n8n

It depends on the scope: how many systems need to be integrated (POS, delivery apps, payments, invoicing), how many locations there are, and how organized the menu and inventory are today. That is why StriqTech does not work with fixed plans. First we understand the need in an initial, no-commitment conversation. Then we put together a proposal with scope, deliverables, dependencies, and acceptance criteria. Only then do we build, test with the people running the restaurant, and agree on launch and maintenance. If the project needs more than workflows, such as a dedicated integration with your POS or a multi-location inventory dashboard, it goes into the same proposal as custom software.

The proposal details which maintenance, hosting, or third-party services are included, and timelines are set phase by phase, separating our work from what depends on the restaurant and on each provider. What you can estimate right now is the third-party infrastructure: n8n Community on a USD 10-15/month VPS, plus the cost of the WhatsApp API and the licenses you already pay for.

5 common mistakes when automating restaurants (and how to avoid them)

1. Syncing inventory without a safety margin. If you show real stock without a buffer, every simultaneous sale cancels orders. Fix: reserve 1-2 units of buffer per critical product.
2. Sending WhatsApp to an old base without opt-in. Guaranteed number ban. Fix: filter customers from the last 90 days and with explicit consent.
3. Not having a manual fallback. If n8n goes down at 11pm on a Friday, you're out of reminders. Fix: health alerts + documented manual protocol.
4. Automating before standardizing. If your menu changes weekly without a process, no workflow will work. Fix: order the process on paper first, then automate it.
5. Throwing an LLM at everything. GPT-5 to decide if there's stock is expensive and slow. Fix: LLM only to understand the customer's message; business logic in deterministic nodes.

Related internal links

• Pillar guide: n8n for Argentine SMBs.
• Related cluster: Evolution API WhatsApp self-hosted.
• Full services: StriqTech process automation.
• Contact: WhatsApp +54 9 11 5499-7296.

CTA

If you have a restaurant billing more than USD 20k/month and you're still closing the register by hand, tell us which process you want to solve: register closing, inventory between the POS and the delivery apps, bookings, or collections. Write to us at +54 9 11 5499-7296 and, after that first conversation, we'll propose the next step with scope, phases, and timelines.