Your First Automated Workflow: Email → Spreadsheet → Notification

Theory is good, but nothing beats building something yourself. In this chapter, we'll use the Make platform (which has the most intuitive interface) as an example to build a real, usable automated workflow from scratch.

Scenario: Automating Customer Inquiry Processing

You have a contact form on your website (or receive customer inquiries via email). Whenever a new inquiry comes in, you need to: 1. Log the inquiry content to Google Sheets 2. Use AI to analyze the inquiry type and urgency level 3. Assign it to the appropriate colleague based on the type 4. Send a notification in a Feishu/DingTalk group.

Doing this process manually takes 5-10 minutes each time. If you receive 20 inquiries a day, that's 50-100 hours of repetitive work per month. Let's solve it with automation.

Build Steps

Step 1: Register for Make and Create a Scenario

Visit make.com to register for a free account. Click "Create a new scenario" to create your first automation flow. Each automation flow in Make is called a Scenario.

Step 2: Set Up the Trigger

The Trigger determines when the automation flow starts. In our scenario, the trigger is "Receiving a new email." Click the plus sign, search for the Email/Gmail module, and select "Watch Emails." Connect your email account and set filter conditions (e.g., specific sender or subject keywords).

Step 3: Add an AI Analysis Node

Add an OpenAI (or other AI) module after the trigger. After configuring the API key, set the Prompt: "Analyze the following customer inquiry and return JSON format: {"type": "pre-sales/after-sales/complaint/other", "urgency": "high/medium/low", "summary": "one-sentence summary"}. Inquiry content: [email body]."

Step 4: Write to Spreadsheet

Add a Google Sheets module and select the "Add a Row" action. Connect your Google account, select the target spreadsheet and worksheet. Map the email information (sender, time, subject) and AI analysis results (type, urgency, summary) to the corresponding columns.

Step 5: Conditional Routing

Add a Router module to set up different processing paths based on the AI-analyzed type. For example: Pre-sales inquiry → Notify sales group, After-sales issue → Notify customer service group, Complaint → Notify manager. Connect each path to a Feishu/DingTalk message-sending module.

Step 6: Test and Enable

Click "Run once" to test the entire flow. Send a test email and check if each node works correctly. Once confirmed, turn on the Scheduling switch and set the check frequency (e.g., check for new emails every 5 minutes).

Core Concepts Explained

Trigger

The starting point of automation—"When something happens." Common triggers: receiving a new email, new data in a spreadsheet, scheduled execution (e.g., daily at 9 AM), receiving a request via Webhook.

Action

What to do after being triggered—"Execute what action." Common actions: sending messages, writing data, calling an API, generating documents.

Router

Decides which path to take based on conditions. Similar to if-else in programming, but configured via a visual interface.

Data Mapping

Passing the output from the previous step as input to the next step. For example, passing the email body to AI analysis, then passing the analysis results to the spreadsheet and notification modules.

Common Issues and Debugging

The most common issue in automation flows is data format mismatch. For example, failure to parse JSON returned by the AI, inconsistent date formats, or errors caused by special characters in text. Make provides detailed error logs; you can view the input/output data for each node individually, making it easy to troubleshoot step by step.

After building your first automation flow, in the next chapter we'll look at enterprise-level automation cases—understanding how real companies use automation to save an entire team's workload.