WhatsApp Business Chatbot with n8n: complete guide for Italian SMBs 2026

Reading time: 9 minutes

In Italy, over 35 million people use WhatsApp every day. Non e’ un social network: e’ il luogo dove i tuoi clienti parlano, decidono e comprano. Eppure, la maggior parte delle PMI italiane continua a gestire WhatsApp come un centralino anni ’90, risposte manuali, tempi morti, opportunita’ perse.

Here's the point: in 2026, a WhatsApp Business chatbot with n8n is no longer a luxury reserved for big companies. It’s an accessible tool, configurable in just a few hours and capable of transforming the way a small business manages customers, orders and leads. In this complete guide I’ll show you exactly how to build one, step by step, with workflow pronti that you can adapt to your own business.

Why a WhatsApp chatbot is essential in 2026

The numbers are clear. According to Meta data updated to 2026, WhatsApp Business messages have a 98% open rate, versus 20-25% for email. The response rate exceeds 45%. No other channel has these metrics in Italy.

But the real paradigm shift is something else: customers expect immediate responses. A 2025 Salesforce study reveals that 73% of Italian consumers abbandona un acquisto se non riceve risposta entro 5 minuti. Cinque minuti. Se sei un artigiano, un ristoratore, un professionista, non puoi permetterti di stare incollato al telefono tutto il giorno.

Un chatbot WhatsApp solves exactly this problem. It responds 24/7, qualifies leads, takes bookings, sends order confirmations and does it all without you having to hire staff. With n8n as the automation engine, hai il controllo totale sui flussi, senza vincoli di piattaforma e senza costi mensili per singola automazione.

WhatsApp Business API: what you need to get started

Before touching n8n, you need access to the WhatsApp Business API. Note: it’s not the same thing as the WhatsApp Business app you find in the store. The APIs are the enterprise layer that enables automation, chatbots and integrations.

Requisiti tecnici

• Meta Business Account verified (business.facebook.com)
• Numero di telefono dedicato: non puo’ essere lo stesso dell’app WhatsApp personale
• BSP (Business Solution Provider): il ponte tra te e le API Meta. I piu’ usati dSMBs italiane: 360dialog, Twilio, WATI
• Istanza n8n: self-hosted o n8n Cloud (consiglio self-hosted su Railway o Hetzner per controllo e costi)

Real costs in 2026

Meta applies a conversation-based pricing. In Italy, the updated costs are:

• Service conversations (replies to customer messages): free for the first 1,000/month, then about €0.03 each
• Conversazioni of marketing (messaggi proattivi, promozioni): circa 0,06 € ciascuna
• BSP: da 0 € (360dialog base) a 49 €/mese (WATI starter)
• n8n self-hosted: 5-10 €/mese su Railway o VPS

For an SMB with 500 conversations per month, the total cost is around 20-60 €/mese. Compare it to a part-time employee handling the same requests: the savings are obvious.

Setting up n8n for WhatsApp: a step-by-step guide

Let's get practical. This guide n8n WhatsApp assumes you already have a working n8n instance and an active BSP account. If you still need to choose your automation platform, first read the confronto n8n vs Make vs Zapier.

Step 1: Configura il webhook di ricezione

In n8n, create a new workflow and add a Webhook. Set it as POST and copy the generated URL. Go into your BSP’s panel (e.g. 360dialog) and paste the URL into the Webhook URL. From this moment on, every incoming WhatsApp message will trigger your workflow.

Step 2: Parsing the message

Aggiungi a nodo Function per estrarre i dati utili dal payload: numero del mittente, testo del messaggio, timestamp, eventuale media allegato. Il payload WhatsApp Cloud API segue una struttura standard JSON, il campo chiave e’ entry[0].changes[0].value.messages[0].

Step 3: Router logico

Usa a nodo Switch to route the message based on its content. A practical example for a restaurant:

• Contiene “menu” o “piatti” → Invia menu del giorno
• Contains “book” or “table” → Start booking flow
• Contains “hours” → Reply with the opening hours
• Other → Reply with a generic message + options

Step 4: Risposte automatiche

For each branch of the router, add a HTTP Request che invia la risposta tramite le API del BSP. Includi pulsanti interattivi (WhatsApp li supporta nativamente) per guidare l’utente in un flusso strutturato. I template di messaggio devono essere approvati da Meta, prevedi 24-48 ore per l’approvazione.

Step 5: Saving data

Ogni conversazione va tracciata. Aggiungi un nodo per salvare i dati su un database (Supabase, Airtable, Google Sheets). Questo ti permette di analizzare i pattern, capire le domande piu’ frequenti e ottimizzare il chatbot nel tempo. Se usi gia’ un CRM, puoi collegarlo direttamente, n8n ha nodi nativi per HubSpot, Pipedrive e decine di altri.

Integrating AI into your WhatsApp chatbot

The real leap in quality comes when you add a layer of artificial intelligence to the chatbot. We’re not talking about canned replies, but a system that understands the customer’s intent and responds naturally.

How it works in practice

In the n8n workflow, between the parsing node and the router, insert a call to an LLM (Claude, GPT-4o, or an open-source model via OpenRouter). The system prompt contains the information about your business: product catalog, return policies, hours, FAQs. The model receives the customer’s message and generates a contextualized response.

This architecture, which I explore in depth in the guide on AI agents for lead generation with n8n, lets the chatbot handle questions you hadn’t anticipated, in natural Italian, with the tone of your business.

Essential guardrails

• Fallback umano: if the chatbot isn’t sure (low confidence score), hand the conversation off to a human agent with a notification
• Limiti of scope: the bot must not make up information about prices or availability. Connect it to real data sources
• GDPR: inform the user that they’re talking to an automated system. It’s a legal requirement in Italy

5 ready-to-use WhatsApp workflows for SMBs

Ecco cinque workflow WhatsApp automation that you can implement with n8n this week. Each workflow builds on the webhook + router foundation described above.

1. Risponditore automatico intelligente

For those who: any SMB. It answers FAQs (hours, location, services) 24/7 with AI. If the question falls outside its scope, it notifies the owner via Telegram or email. Setup time: 2 hours.

2. Booking appuntamenti

For those who: professional practices, hairdressers, restaurants. The customer writes “I want to book,” the bot proposes the available slots (connected to Google Calendar), confirms and sends a reminder 24h before. Setup time: 3 hours.

3. Conferma ordine + tracking

For those who: e-commerce, artisans with direct sales. After the purchase (triggered by WooCommerce or Stripe), the customer receives confirmation on WhatsApp, then automatic tracking when the package is shipped. Integrate with email marketing for a complete follow-up.

4. Scored lead qualification

For those who: agencies, consultants, B2B. The bot asks 3-4 structured questions (budget, timeline, industry), assigns a score, and routes hot leads straight to your calendar. Cold leads enter a nurturing sequence. Learn more with the guide to automated lead scoring with n8n.

5. Post-sale survey with NPS

For those who: any SMB that wants to improve. 48 hours after the purchase or the service, the bot sends a short survey (3 questions + NPS). The results land in a Notion dashboard or Google Sheets. Dissatisfied customers receive an immediate follow-up from the owner.

ROI: how much you save with a WhatsApp chatbot

Parliamo di numeri concreti. Prendiamo una PMI italiana tipica, uno studio dentistico a Verona con 30-40 chiamate al giorno per prenotazioni e informazioni.

Before the chatbot

• 1 receptionist part-time dedicata al telefono: ~1.200 €/mese
• Chiamate perse fuori orario: ~15/giorno (potenziali pazienti persi)
• Average time to phone booking: 4 minuti
• No-shows without reminders: 18% of appointments

Dopo the chatbot

• Costo chatbot (BSP + n8n + AI): ~80 €/mese
• 70% of inquiries handled automatically
• Zero chiamate perse, il bot risponde anche alle 3 di notte
• Booking time via bot: 45 secondi
• No-shows with WhatsApp reminders: sceso al 6%

Risparmio netto stimato: 900-1.100 €/mese, not counting the value of leads recovered after hours. ROI is reached as early as the first month.

E non e’ solo questione di soldi. La receptionist, liberata dalle chiamate ripetitive, puo’ dedicarsi all’accoglienza in studio, un’attivita’ ad alto valore che nessun bot puo’ sostituire.

Approfondimenti consigliati

This guide is part of a complete series on automation for Italian SMBs. If you want to go deeper:

• n8n vs Make vs Zapier: which to choose in 2026: il confronto definitivo per PMI italiane
• Automated lead scoring with n8n and HubSpot: qualifica i tuoi contatti senza muovere un dito
• Automated email marketing with n8n: workflow pronti per newsletter, follow-up e nurturing
• AI Agent for lead generation with n8n and Claude AI: il livello successivo dell’automazione intelligente

Prossimi passi

A WhatsApp Business chatbot with n8n is not a months-long project. With the right foundations, you can have a working system in a week. But the difference between a bot that works and one that delivers results lies in the strategy: which flows to automate, how to integrate AI, where to put the guardrails.

Se vuoi implementare un a custom WhatsApp chatbot for your SMB: o vuoi capire quale dei 5 workflow ha piu’ senso per il tuo caso specifico, let's talk about it in a free 30-minute call. I'll analyze your situation and show you exactly what to automate first.

Mattia Calastri is an AI and marketing automation consultant in Verona. He helps Italian SMBs and professionals automate their commercial processes with n8n, WhatsApp Business API and artificial intelligence. He writes on mattiacalastri.com.