WhatsApp + Google Sheets: support replies, logged
Your WhatsApp inbox is where good support goes to die. Same questions, same answers, and somehow you still end up rewriting them at midnight because nobody can find the “right” reply.
This WhatsApp FAQ logging setup hits support leads first, but solo founders and ops managers feel it too. You get consistent, on-brand replies pulled from your own Google Doc, and every conversation gets logged to Google Sheets so you can finally see what customers ask most.
Below is the workflow, what it automates, and what changes once you stop answering everything by hand.
How This Automation Works
The full n8n workflow, from trigger to final output:
n8n Workflow Template: WhatsApp + Google Sheets: support replies, logged
flowchart LR
subgraph sg0["AI Agent Flow"]
direction LR
n0@{ icon: "mdi:robot", form: "rounded", label: "AI Agent", pos: "b", h: 48 }
n1["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/code.svg' width='40' height='40' /></div><br/>cleanAnswer"]
n2["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/code.svg' width='40' height='40' /></div><br/>24-hour window check"]
n3@{ icon: "mdi:swap-horizontal", form: "rounded", label: "If", pos: "b", h: 48 }
n4@{ icon: "mdi:brain", form: "rounded", label: "Google Gemini Chat Model", pos: "b", h: 48 }
n5["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/aitransform.svg' width='40' height='40' /></div><br/>Prepare Prompt"]
n6@{ icon: "mdi:database", form: "rounded", label: "Google Sheets", pos: "b", h: 48 }
n7@{ icon: "mdi:cog", form: "rounded", label: "Date & Time", pos: "b", h: 48 }
n8@{ icon: "mdi:cog", form: "rounded", label: "company's knowledge", pos: "b", h: 48 }
n9@{ icon: "mdi:memory", form: "rounded", label: "Simple Memory", pos: "b", h: 48 }
n10["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/webhook.dark.svg' width='40' height='40' /></div><br/>WhapAround - Listeing"]
n11["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/webhook.dark.svg' width='40' height='40' /></div><br/>WhapAround - Respond Message "]
n12["<div style='background:#f5f5f5;padding:10px;border-radius:8px;display:inline-block;border:1px solid #e0e0e0'><img src='https://flowpast.com/wp-content/uploads/n8n-workflow-icons/webhook.dark.svg' width='40' height='40' /></div><br/>WhapAround - Respond Template "]
n3 --> n1
n3 --> n12
n0 --> n7
n7 --> n6
n1 --> n11
n6 --> n2
n9 -.-> n0
n5 --> n0
n8 --> n5
n2 --> n3
n10 --> n8
n4 -.-> n0
end
%% Styling
classDef trigger fill:#e8f5e9,stroke:#388e3c,stroke-width:2px
classDef ai fill:#e3f2fd,stroke:#1976d2,stroke-width:2px
classDef aiModel fill:#e8eaf6,stroke:#3f51b5,stroke-width:2px
classDef decision fill:#fff8e1,stroke:#f9a825,stroke-width:2px
classDef database fill:#fce4ec,stroke:#c2185b,stroke-width:2px
classDef api fill:#fff3e0,stroke:#e65100,stroke-width:2px
classDef code fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
classDef disabled stroke-dasharray: 5 5,opacity: 0.5
class n0 ai
class n4 aiModel
class n9 ai
class n3 decision
class n6 database
class n10,n11,n12 api
class n1,n2 code
classDef customIcon fill:none,stroke:none
class n1,n2,n5,n10,n11,n12 customIcon
The Problem: WhatsApp Support Becomes Repetitive (and Untrackable)
WhatsApp support looks simple until it’s not. A customer asks about pricing, delivery, refunds, opening hours, booking details, or “can I speak to a human?” and you scramble to answer fast so they don’t disappear. Then another message comes in, and another. Before you know it, your “quick replies” are outdated, different teammates answer the same question in different ways, and you’ve got no clean record of what’s being asked. The worst part is the invisible cost: you can’t improve what you can’t measure.
It adds up fast. Here’s where it breaks down.
- Teams keep retyping the same FAQ answers, which quietly burns a couple hours every week.
- Replies drift over time, so pricing, policies, or tone end up inconsistent across agents.
- Without a proper log, you’re guessing which issues drive the most tickets and refunds.
- When WhatsApp’s 24-hour reply window rules kick in, late responses create awkward back-and-forth and missed conversions.
The Solution: AI Replies from Google Docs, Logged to Sheets
This workflow turns your Google Doc into your living support knowledge base, then uses it to answer WhatsApp messages automatically. When a new WhatsApp message comes in through a webhook, the workflow pulls the latest content from your Google Docs document (so you’re not updating prompts in five places). It assembles that knowledge into a clear “answer like this” instruction, sends the customer question to an AI agent (OpenAI or Gemini), and generates a reply that matches your documentation. Next, it stamps the message with the current date/time and logs the whole interaction to Google Sheets for reporting. Finally, it checks the 24-hour window and sends either a normal reply or a WhatsApp template reply when needed.
The workflow starts with an incoming WhatsApp webhook, then reads your Google Doc and builds a prompt for the AI. After the AI drafts the response, n8n cleans up the text, logs the conversation to Google Sheets, and routes the final message to the correct WhatsApp response type.
What You Get: Automation vs. Results
| What This Workflow Automates | Results You’ll Get |
|---|---|
|
|
Example: What This Looks Like
Say you get 20 WhatsApp questions a day and about half are repeats (hours, delivery, refund policy). Manually, even “quick” answers take maybe 3 minutes each once you read, copy, tweak, and send, so that’s about 30 minutes daily just on repeats. With this workflow, you spend roughly 10 minutes updating your Google Doc when something changes, then replies and logging run automatically in the background. Over a week, most teams get a few hours back and the spreadsheet shows exactly what to fix next.
What You’ll Need
- n8n instance (try n8n Cloud free)
- Self-hosting option if you prefer (Hostinger works well)
- WhapAround to connect your WhatsApp number and webhook.
- Google Docs for your FAQ and business knowledge base.
- OpenAI API key or Gemini API access (get it from your OpenAI or Google AI console)
Skill level: Beginner. You’ll paste a doc ID, connect accounts, and map a few message fields.
Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).
How It Works
A WhatsApp message triggers the workflow. WhapAround forwards the inbound chat to n8n through the “Incoming WhatsApp Hook,” including who sent it and the message text.
Your Google Doc is fetched as the source of truth. n8n pulls the document you specify (“Retrieve Knowledge Doc”), so your policies and FAQs live in one place and stay up to date without editing the automation.
The AI drafts a reply with memory when available. The workflow assembles a clean prompt, sends it to the AI agent (OpenAI or Gemini), and can keep short chat context using a session memory buffer so replies feel less robotic in longer conversations.
Everything gets logged, then routed through the right reply method. The workflow stamps the current time, writes a row to Google Sheets, checks the WhatsApp 24-hour window, and sends either a standard reply or a template reply.
You can easily modify the Google Sheet columns to match your reporting, or swap the AI engine from OpenAI to Gemini based on your preference. See the full implementation guide below for customization options.
Step-by-Step Implementation Guide
Step 1: Configure the Webhook Trigger
Set up the incoming WhatsApp webhook that starts the workflow and hands off the payload for processing.
- Add and open Incoming WhatsApp Hook.
- Set HTTP Method to
POST. - Set Path to
30720c7c-18f4-4815-be3b-03343d53ee45. - Set Response Mode to
responseNodeso the response is handled by Send WhatsApp Reply or Send Template Reply. - Connect your WhatsApp provider to the webhook URL generated by Incoming WhatsApp Hook.
⚠️ Common Pitfall: The code in Assemble Prompt Text and Validate 24h Window references when message received. If your webhook node is named Incoming WhatsApp Hook, update those references or rename the webhook node to match to avoid runtime errors.
Step 2: Connect Google Docs for Knowledge Retrieval
Pull your knowledge base content from a Google Doc so the assistant can answer questions using current support materials.
- Add and open Retrieve Knowledge Doc.
- Set Operation to
get. - Set Document URL to
[YOUR_ID](replace with your Google Doc URL or ID). - Credential Required: Connect your Google Docs credentials.
Step 3: Assemble the AI Prompt and Configure the Agent
Create a prompt that combines the date, knowledge document text, and the incoming WhatsApp message, then send it to the AI agent.
- Open Assemble Prompt Text and keep the provided jsCode for building the
finalPrompt. - Verify the prompt format in Assemble Prompt Text matches the expected structure in the instructions.
- Open Support AI Responder and set Text to
{{ $json.finalPrompt }}. - In Support AI Responder, review the System Message to ensure it matches your support voice and policies.
- Open Gemini Chat Engine and set Model Name to
models/gemini-2.5-flash-preview-04-17-thinking. - Credential Required: Connect your Google Gemini credentials in Gemini Chat Engine. This is the language model used by Support AI Responder.
- Open Session Memory Store and set Session Key to
{{ $('when message received').item.json.contacts[0].wa_id }}and Session ID Type tocustomKey.
Session memory is attached to Support AI Responder via Session Memory Store. Add any necessary credentials to the parent model node (Gemini Chat Engine), not the memory sub-node.
Step 4: Log Activity and Validate the 24-Hour Window
Log interactions to a spreadsheet and determine whether the message falls within the 24-hour WhatsApp response window.
- Open Current Date Time and keep default settings to capture the timestamp after AI processing.
- Open Log to Spreadsheet and configure the target spreadsheet and sheet where you want to store conversation logs.
- Credential Required: Connect your Google Sheets credentials.
- Open Validate 24h Window and keep the provided jsCode to calculate
withinWindowfrom the WhatsApp timestamp.
Step 5: Route Replies Based on the 24-Hour Decision
Use a decision gate to send an immediate AI reply within 24 hours or a template fallback outside the window.
- Open Decision Gate and confirm the condition uses
{{ $json.withinWindow }}with the boolean operator set to true. - On the “true” output, keep the connection to Polish Reply Text to clean the AI output.
- On the “false” output, keep the connection to Send Template Reply to send a fallback response.
- Open Polish Reply Text and keep the provided jsCode to strip formatting and unwanted preambles.
Step 6: Configure WhatsApp Responses
Return the AI response or a template message back to the original webhook request.
- Open Send WhatsApp Reply and format the response body to match your WhatsApp provider’s expected response payload.
- Open Send Template Reply and configure the fallback response payload for messages outside the 24-hour window.
Step 7: Test & Activate
Run a full test to confirm the webhook, AI reply, logging, and routing behave as expected, then activate the workflow.
- Click Execute Workflow and send a test WhatsApp message to the webhook URL from Incoming WhatsApp Hook.
- Verify Retrieve Knowledge Doc returns your document content and Assemble Prompt Text outputs
finalPrompt. - Confirm Support AI Responder produces a response and Polish Reply Text returns cleaned text.
- Check that Log to Spreadsheet writes a new row in your Google Sheet.
- Ensure Decision Gate routes to Send WhatsApp Reply within 24 hours and to Send Template Reply otherwise.
- When satisfied, toggle the workflow to Active for production use.
Common Gotchas
- Google Docs access can fail if the document isn’t shared correctly or the connected Google account loses permissions. If the workflow suddenly replies with “missing context,” check the Google connection and the doc sharing settings first.
- If you’re using Wait nodes or external rendering, processing times vary. Bump up the wait duration if downstream nodes fail on empty responses.
- OpenAI or Gemini outputs can be bland with default prompts. Add your brand voice and “what not to say” rules early, or you will be editing responses forever.
Frequently Asked Questions
About 15-30 minutes if your accounts are ready.
No. You’ll mainly connect services and map message fields. The only “technical” part is pasting your Google Doc ID and API key in the right spots.
Yes. n8n has a free self-hosted option and a free trial on n8n Cloud. Cloud plans start at $20/month for higher volume. You’ll also need to factor in OpenAI or Gemini API usage, which is usually small for short FAQ replies.
Two options: n8n Cloud (managed, easiest setup) or self-hosting on a VPS. For self-hosting, Hostinger VPS is affordable and handles n8n well. Self-hosting gives you unlimited executions but requires basic server management.
Yes, and you should, honestly. A common pattern is to add an “If” rule after the AI response that looks for certain intents (refund, cancel, legal, angry messages) and then routes the chat to a human via a Telegram alert. You can also tweak the “Assemble Prompt Text” step to include your escalation policy and business hours, so the assistant knows when to say “A teammate will reply tomorrow.” Many teams also customize the Google Sheets log to capture a category column, which makes weekly reporting much easier.
Usually it’s permissions or an expired Google credential. Make sure the connected Google account can actually open the document, then re-authenticate Google in n8n and confirm the document ID is correct. If the doc was moved between Drives or the sharing settings changed, the workflow can suddenly start pulling empty content, which makes the AI responses worse right away.
On self-hosted n8n, it mainly depends on your server and your AI rate limits.
Often, yes. n8n is more comfortable when you need branching logic like “log it, check the 24-hour rule, then choose a reply path,” and it’s easier to add memory and custom processing without hitting premium limits. Self-hosting also matters if you have busy periods and don’t want to pay per task. Zapier or Make can be faster to click together for a two-step flow, but support automation tends to grow extra rules over time. Talk to an automation expert if you want help choosing.
Once this is running, your WhatsApp replies stay consistent and your spreadsheet becomes a simple “top issues” dashboard. Set it up once, then get on with real work.
Need Help Setting This Up?
Our automation experts can build and customize this workflow for your specific needs. Free 15-minute consultation—no commitment required.
