How to automate WhatsApp order processing for small businesses
Learn how to automate WhatsApp order processing — from menu to order confirmation and Google Sheets logging — for bakeries, restaurants, and small retailers.
Rebecca Pearson6 min read
For many small businesses — bakeries, food businesses, local retailers — WhatsApp has become the de facto ordering channel. Customers message to ask what's available, you respond manually, they confirm their order, and you write it down somewhere. It works, but it doesn't scale well. Every order is manual. Every response takes time. And when it's busy, things fall through the cracks.
WhatsApp order processing automation changes this. Your bot handles the whole flow — presenting options, taking the order, calculating the total, confirming with the customer, and logging it — without anyone on your team having to touch it.
TL;DR
- The full order flow — from customer message to Google Sheets log and confirmation — can be automated completely with CodeWords and Cody, the AI automation assistant.
- Natural language ordering means customers don't need to navigate a rigid menu system — they can say "two sourdoughs and a croissant for Thursday" and the bot understands.
- Bakeries, restaurants, and small retailers are the ideal fit — anywhere orders are simple, repetitive, and volume-driven.
Why WhatsApp is the natural ordering channel for small businesses
Small businesses don't always have the resources to build or maintain a custom ordering platform. WhatsApp is a pragmatic alternative — customers already have it, it's free to use, and it supports the kind of informal, conversational ordering that small businesses thrive on.
The problem is that manual WhatsApp ordering doesn't scale. At five orders a day, it's manageable. At 50, it's a full-time job. Automation lets you serve the same volume with significantly less effort — or grow volume without hiring.
The automated order flow
Here's how a fully automated WhatsApp order processing flow works:
1. Customer messages in: the customer sends a message — "I'd like to place an order" or simply "Do you have sourdough available Thursday?" — and the bot responds immediately.
2. Bot presents the menu or catalog: the bot shares what's available, either as a formatted text list or by pulling live data from Google Sheets or Notion. If stock or availability changes daily (common for bakeries), the bot can pull the current day's menu from a spreadsheet that you update each morning.
3. Customer orders in natural language: this is where a well-designed bot pays off. Instead of requiring customers to navigate numbered options, the bot lets them say what they want naturally: "I'll take two of the sourdough, one croissant, and a seeded rye please." The bot parses this and confirms back what it's understood.
4. Bot confirms the total and collection details: once the order is understood, the bot confirms: "That's two sourdough loaves, one croissant, and one seeded rye — total £14.50. Collection Thursday anytime from 8am. Shall I confirm this order?"
5. Customer confirms: "Yes please." The bot logs the order and sends a final confirmation.
6. Order logs to Google Sheets: automatically, the order is written to a Google Sheets row — customer name, phone number, items, total, collection date, timestamp. Your team can see all orders in one place without any manual data entry.
7. Confirmation message to the customer: the customer receives a neat confirmation they can refer back to: "Your order is confirmed for Thursday. Order ref: #247. See you then."
The whole exchange takes two to three minutes. Your team's involvement is zero until they're fulfilling the order.
Building the flow with CodeWords
Here's how to set this up:
Step 1: Sign up at CodeWords and open a new agent. Describe your order processing flow to Cody — include what you sell, how orders are structured, and where you want them logged.
A good starting description:
"Build a WhatsApp order bot for a bakery. When a customer messages, it should greet them and share today's menu (pulled from a Google Sheet). It should take their order in natural language, confirm back what they've ordered with a total, ask for their collection date, and log the confirmed order to a separate Google Sheet with their name, number, items, total, and date."
Step 2: Connect WhatsApp. Personal Device is the simplest option for most small businesses — it takes 30 seconds via pairing code.
Step 3: Connect Google Sheets via Composio. You'll need two sheets: one for your menu/stock (which the bot reads), and one for confirmed orders (which the bot writes to).
Step 4: Set up your menu sheet. A simple structure works well — columns for item name, description, price, and availability (yes/no). Update the availability column each morning to reflect what you have that day.
Step 5: Test the flow end to end. Place a test order, check that the confirmation reads naturally, verify the Google Sheets entry is correct.
Step 6: Go live. Share your WhatsApp number via your website, social media, and any packaging inserts.
Handling common edge cases
Out-of-stock items: if a customer orders something that's marked as unavailable in your sheet, the bot should say so gracefully and suggest alternatives: "We're actually out of seeded rye this week — would sourdough work instead?"
Custom requests: "Can I get the croissants with a bit of butter on the side?" Your bot should acknowledge this and log it as a note on the order, rather than getting confused.
Order changes: "Actually, can I change it to one sourdough instead of two?" The bot should be able to update an order before it's been confirmed, then re-confirm the revised total.
Pricing questions: "How much is the sourdough?" A simple lookup in the menu sheet, answered immediately.
These edge cases are handled naturally by a well-prompted AI bot — you don't need to write explicit rules for each one. The system prompt guides the model on how to behave, and it handles variations on its own.
Who this is built for
Bakeries: this is the most natural fit. Daily menus, limited stock, regular customers who already message to order. See the bakery WhatsApp agent guide for more.
Small food businesses and delis: similar dynamics to bakeries — daily specials, fresh stock, high volume of WhatsApp enquiries.
Small retailers with a regular product catalog: if you have a defined set of products and customers regularly message to enquire and order, automation handles the repetitive parts cleanly.
Businesses that take made-to-order requests: custom cakes, personalised gifts, print-on-demand products — anywhere the order requires some back-and-forth before it's finalised.
Connecting to Stripe for payment (optional)
For businesses that want to take payment at point of order, CodeWords integrates with Stripe via Composio. Once an order is confirmed, the bot can generate a payment link and send it to the customer — "To complete your order, here's your payment link: [link]. Payment is required by Wednesday evening."
The bot can check payment status before finalising the order and log the payment reference alongside the order in your Google Sheet.
This isn't always necessary — many small businesses prefer to take payment on collection — but it's a straightforward add-on if cash flow or no-shows are a concern.
The impact on your business
Small businesses that automate WhatsApp order processing typically see:
- Fewer missed orders (the bot is available 24/7, including evenings and weekends)
- Faster order processing (customers get an immediate response rather than waiting for you to be free)
- Cleaner order records (Google Sheets log replaces handwritten notes or WhatsApp screenshots)
- More capacity to handle volume growth without additional staff overhead
It's not about replacing the personal touch of a small business — it's about removing the mechanical, repetitive parts of order-taking so you can focus on the parts that matter.
Start building your WhatsApp order processing bot on CodeWords and describe what you need to Cody.