Auto-Draft Email Replies to Form Submissions

Imagine every time someone submits your feedback form, a personalized draft reply is already waiting in your Gmail inbox, ready for your review. That is exactly what you can build with Google Workspace Studio, Google’s new no-code automation platform.

In this tutorial, I’ll walk you through a real-world Workspace Studio workflow that watches for Google Form submissions, uses Gemini to decide whether a response is needed, extracts key details, drafts an email reply, and sends you a notification in Google Chat. The entire setup takes about 15 minutes, and you will not write a single line of code. If you have ever built automations with Apps Script, you will appreciate just how much faster this approach is for straightforward Workspace-to-Workspace workflows.

Prerequisites

This tutorial assumes the following prerequisites:

• A Google Workspace account (Business Starter or higher) with Workspace Studio enabled by your admin.
• A Google Form that collects feedback or questions, including an email address field.
• Basic familiarity with Google Forms, Gmail, and Google Chat.

What is Google Workspace Studio?

Before we dive into the case study, let me give you a quick overview of Workspace Studio. Launched by Google in late 2025, Workspace Studio is a no-code platform for building AI-powered automations, which Google calls “agents”, that work across Gmail, Google Drive, Google Sheets, Google Chat, Google Forms, and other Workspace apps.

The key idea behind Workspace Studio is that you describe your automation in plain language, and Gemini builds the workflow for you. You can also construct workflows manually step by step using a visual editor. Each workflow consists of a Starter (the trigger), followed by one or more Actions like asking Gemini a question, extracting data, drafting an email, or sending a Chat notification. Conditional logic is handled through Decide and Check if steps, which let Gemini evaluate conditions using natural language rather than rigid if-then rules.

You can access Workspace Studio at studio.workspace.google.com or by clicking the double-arrow icon in the top-right corner of any Workspace web app.

Workspace Studio vs. Apps Script

If you are a regular reader of this blog, you are probably wondering how Workspace Studio compares to Apps Script. Both tools automate Google Workspace, but they serve different needs and different users.

Accessibility. Workspace Studio requires zero coding. You build workflows by selecting steps from a visual interface and writing natural-language prompts for Gemini. Apps Script is a JavaScript-based scripting platform that does require programming knowledge, though AI coding assistants like Gemini and Claude can help you write Apps Script code even if you are not a developer.

AI capabilities. Workspace Studio has Gemini built in as a first-class citizen. Steps like Decide, Extract, and Ask Gemini let you inject AI reasoning directly into your workflow without any API configuration. In Apps Script, you would need to call the Gemini API yourself using UrlFetchApp, manage API keys, and parse JSON responses manually. One important note: Gemini-powered steps in Workspace Studio are only available if your Workspace admin has enabled Gemini for your domain. If Gemini is not enabled, you will not see AI steps like Decide, Extract, or Ask Gemini in the editor.

Triggers. Workspace Studio offers a rich set of starters out of the box: scheduled runs, incoming emails, Chat messages, Chat mentions, emoji reactions, sheet changes, Drive file and folder events, calendar meetings, meeting transcripts, and form submissions. Apps Script supports many of these same events through installable triggers, plus time-driven triggers and spreadsheet edit triggers. Both platforms cover the most common automation scenarios.

Flexibility and power. This is where Apps Script still wins. Apps Script gives you full programmatic control: loops, custom functions, complex data transformations, direct access to every Google API, and the ability to call external services. Workspace Studio is intentionally simpler, which means it cannot handle highly custom logic or complex data processing on its own.

The best of both worlds. Google has introduced a limited preview feature that lets you build custom steps in Workspace Studio using Apps Script. This means you can create reusable custom actions, written in Apps Script, and plug them into any Workspace Studio workflow. This is a hybrid approach: you get the visual, no-code workflow builder of Studio for orchestration, with the full flexibility of Apps Script for any step that needs custom logic. If this feature is available in your domain, it is worth exploring.

Best practice. Use Workspace Studio when your workflow is primarily about connecting Workspace apps together with AI reasoning in the middle. Reach for Apps Script when you need fine-grained control, complex logic, or integration with non-Workspace APIs. And when you need both, use custom Apps Script steps inside Studio to get the best of both worlds.

7 steps to auto-draft email replies

Here is an overview of the workflow we are building. Each step maps directly to a step in the Workspace Studio visual editor.

• Step 1 — Set up the form trigger
• Step 2 — Use Decide to check for an email address
• Step 3 — Add a Check if condition
• Step 4 — Extract the recipient email and subject
• Step 5 — Ask Gemini to draft a reply
• Step 6 — Draft the email in Gmail
• Step 7 — Notify yourself in Google Chat

Step 1 — Set up the form trigger

Open Workspace Studio by navigating to studio.workspace.google.com. Click Create to start a new workflow. In the Starter section, select When a form response comes in. Studio will ask you to choose the Google Form you want to monitor. Select your feedback form from the picker.

Once selected, Studio automatically makes all of the form’s question fields available as variables you can use in later steps. In this example, the form I’m using is called “Spreadsheet </> Dev Feedback” and it has fields for a reaction (Helpful, Insightful, Confusing, or Bug/Error) and a free-text feedback field where the reader can optionally provide their email address if they want a response.

Step 2 — Use Decide to check for an email address

Not every form submission will include an email address. Before we try to draft a reply, we need to determine whether there is actually someone to reply to and if the feedback submission requires a response. This is where the Decide step comes in.

Click Add step and select Decide. The Decide step asks Gemini to evaluate a condition and return a true or false result. In the prompt field, I entered the following:

Does [Step 1: Your feedback or question] contain an email
address? If yes, does it require a response: Yes or No?

Notice how the prompt references a variable from Step 1 using the + Variables button. This is how data flows between steps in Workspace Studio: by inserting variable chips that pull values from earlier steps. Gemini reads the submission content and makes a judgment call about whether it contains an email and warrants a response.

This is fundamentally different from Apps Script, where you would write a regular expression to detect an email pattern. Here, Gemini uses natural language understanding, which means it can handle edge cases like emails embedded in sentences or non-standard formats.

Step 3 — Add a Check if condition

The Check if step acts as a gate. It looks at the output of the previous Decide step and only runs the remaining steps if the condition is true. Click Add step and select Check if. Set the condition to: if Step 2: Decision is true.

All subsequent steps are nested inside this Check if block. This means that if Gemini determines the submission does not contain an email or does not need a reply, the workflow stops here.

Step 4 — Extract the recipient email and subject

Now that we know a reply is warranted, we need to pull out structured data from the free-text submission. The Extract step uses Gemini to find specific pieces of information in unstructured content. Click Add step and choose Extract.

In the Content field, insert the variable for the feedback text from Step 1. Then, under What to extract, define two custom fields. The first is Recipient Email with the description “Email of the person who submitted the feedback.” The second is Subject in 50 characters with the description “Extract the summary of the feedback as a subject line.”

Gemini will parse the free-text submission and return these two values as separate variables that you can use in later steps. This is very powerful. In Apps Script, extracting an email from unstructured text would require regex, and generating a subject line summary would require a separate API call to a language model.

Step 5 — Ask Gemini to draft a reply

This is the heart of the workflow. The Ask Gemini step generates free-form text based on a prompt you provide. Click Add step and select Ask Gemini. Here is the prompt I used:

Draft a response to the feedback or question below
submitted by a reader.
 
The person who submitted the feedback is:
[Step 4: Recipient Email]
 
Their reaction about this post was:
[Step 1: Reaction]
 
Feedback:
[Step 1: Your feedback or question]
 
Here are some instructions:
1. Draft in narrative form. Direct, friendly, helpful tone.
2. Be brief.

I also enabled both Web search and Workspace as sources Gemini can use. This means Gemini can look up the actual blog post URL to understand context before drafting the reply. The variable chips (shown in blue in the Studio editor) pull values from Steps 1 and 4 dynamically at runtime.

Step 6 — Draft the email in Gmail

Click Add step and select Draft an email. This step creates a draft in your Gmail inbox. It does not send it automatically, which is an important safety feature. Wire up the fields using variables from previous steps: set To to the Recipient Email from Step 4, set Subject to the Subject in 50 characters from Step 4, and set Message to the Content created by Gemini from Step 5.

When the workflow runs, you will find a fully composed draft in your Gmail Drafts folder. You can review it, make any edits, and send it when you are ready. This human-in-the-loop approach is a best practice for AI-generated emails. You get the speed benefit of automation without the risk of sending something incorrect.

Step 7 — Notify yourself in Google Chat

The final step sends you a heads-up in Google Chat so you know a new draft is waiting. Click Add step and select Notify me in Chat. In the message field, type something like “I’ve drafted an email for you here:” followed by the Email ID variable from Step 6. This gives you a direct link to the draft right inside your Chat notifications.

And that is the entire workflow: seven steps, zero code, and roughly 15 minutes of setup time.

Testing and activating the workflow

Before you turn on the workflow, use the Test run button at the bottom of the Studio editor. Studio will simulate the workflow using a recent form submission or ask you to submit a test response. Review the output at each step to make sure variables are flowing correctly and Gemini’s responses look reasonable.

Once you are satisfied, click Save changes and then Turn on to activate the workflow. From this point forward, every new form submission will be processed automatically. You can monitor activity and past runs from the Studio dashboard.

Conclusion

In this tutorial, I walked you through building a Workspace Studio workflow that auto-drafts email replies to Google Form submissions using Gemini. You learned how to set up a form trigger, use Decide and Check if for conditional logic, extract structured data from free text, prompt Gemini to generate an email draft, and notify yourself via Google Chat.

Workspace Studio is not a replacement for Apps Script, it is a complement. For AI-first, cross-app workflows that live entirely within Google Workspace, Studio is faster to set up and easier to maintain. For complex data processing, custom UIs, or integrations outside the Workspace ecosystem, Apps Script remains the better tool. The best practice is to know both and pick the right one for the job.

Thank you for reading, and I hope this tutorial helps you save time on your next automation project.