We Tested Google's New Workspace Studio on 100 Emails--It Still Can't Keep Spam Out of Your Inbox

Google just launched Workspace Studio--a no-code workflow builder baked right into Gmail, Chat, Drive, and Calendar. Describe what you want in plain English, and Gemini builds an automation for you. Pre-built templates. Drag-and-drop steps. Connectors to Salesforce, Jira, and more.

It sounds like exactly what your inbox needs. It isn't.

We ran Studio's email workflows against 100 real emails--a mix of legitimate messages, AI-generated cold outreach, BDR campaigns, and newsletter spam. The result? Studio got fooled by the same tricks that fool every simple filter. It flagged spam as urgent. It buried real emails. And it proved what we've been saying for months: a workflow is not a spam filter.

What Google Workspace Studio Actually Does

Studio is Google's answer to Zapier and Power Automate, but native to Workspace. You describe a workflow in natural language--"notify me about urgent emails," "save attachments to Drive," "summarize meeting notes in Chat"--and Gemini generates the automation steps.

It ships with pre-built templates like:

• "Notify me about urgent emails" -- Flags messages Gemini thinks are important
• "Summarize new emails" -- Sends digests to Chat
• "Extract action items" -- Pulls tasks from email threads

These are useful for productivity. Summarizing meeting threads? Great. Auto-filing attachments? Handy. But the moment you point Studio at your inbox and ask it to separate signal from noise, it falls apart.

What We Tested

We took 100 emails from a real inbox and ran them through Studio's email workflows. The mix:

• 30 legitimate emails: Client replies, team threads, invoices, shipping notifications
• 35 AI-generated cold outreach: Personalized BDR emails from platforms like Outreach, Apollo, and Salesloft
• 20 automated sales campaigns: Multi-touch sequences with fake reply threads and manufactured urgency
• 15 newsletter/subscription spam: Unwanted signups, promotional blasts, and graymail

We tested two approaches:

1. The "Notify me about urgent emails" template -- Studio's own pre-built workflow, used exactly as shipped
2. A custom workflow -- We described what we wanted: "Flag emails from real people I've communicated with before. Ignore cold outreach and sales pitches."

How Studio Got Tricked

The failures weren't random. Studio fell for the exact tactics that modern cold outreach is designed to exploit.

Fake Urgency Works Every Time

Cold outreach emails are engineered to look urgent. Subject lines like "Quick question about [your company]" and "Following up on our conversation" triggered Studio's urgency detection every time. Studio doesn't know you never had a conversation with that sender. It just sees the word "following up" and flags it.

Personalization Fools Simple Classification

Modern BDR tools pull your name, company, role, and recent activity from LinkedIn and public sources. An email that opens with "Hey Jeff, saw Email Ferret just shipped a new feature--congrats!" reads like a real message to any system that only analyzes content. Studio treated these personalized cold pitches as legitimate business communication.

Warm Domains Pass Every Check

Sales teams warm their sending domains for weeks before launching campaigns. By the time their emails hit your inbox, the domain has a clean reputation. Studio has no domain intelligence layer. If the domain isn't on a blocklist, the email sails through.

Fake Reply Threads Slip Past

Some of the most deceptive cold outreach uses fake "Re:" prefixes and manufactured thread history. Studio's workflows don't verify whether a conversation thread is real. They see "Re: Our discussion" and treat it as an ongoing conversation.

Real Emails Got Buried

This is the worst part. Because Studio flagged so much spam as urgent, the actual urgent emails--a client asking for a deadline change, a billing alert, a teammate flagging a production issue--got lost in the noise. When everything is urgent, nothing is.

Why Simple Workflows Will Never Solve This

Studio's workflows fail at spam detection for a structural reason: they only look at one dimension of an email at a time.

A workflow can check subject lines for keywords. It can look at whether you've emailed the sender before. It can scan body text for sales language. But it does each of these in isolation, as discrete steps in a pipeline.

Effective spam detection requires evaluating dozens of signals simultaneously and weighting them against each other:

• Domain signals: How old is the domain? What's its DNS configuration? Does it match known cold outreach infrastructure?
• Sender behavior: Is this a first-time sender? Are they using automation tool headers? What are their sending patterns?
• Content analysis: Does the language match sales outreach patterns? Is the personalization genuine or templated?
• Thread context: Is this a real reply or a manufactured thread? Is there prior engagement?
• Trust signals: Has this sender been allowlisted? Are they from a known organization?

Studio workflows can't do this. They're linear automations--if X then Y. Spam detection requires scoring across all of these dimensions simultaneously and making a holistic judgment. That's a fundamentally different kind of system.

What Actually Keeps Spam Out of Your Inbox

Stopping modern cold outreach and AI-generated spam requires purpose-built detection that goes far beyond what a no-code workflow can do.

Multi-Signal Heuristic Scoring

Effective filters score every email across dozens of weighted signals simultaneously. A single suspicious indicator might not mean anything. But a new domain + sales language + automation headers + no prior contact + fake personalization? That combination tells the full story. Heuristic analysis catches what keyword matching misses.

Automation Fingerprinting

Cold outreach platforms leave fingerprints in email headers, sending patterns, and infrastructure. Dedicated detection engines know what Outreach, Apollo, Salesloft, and Instantly look like under the hood. Studio doesn't check headers at all.

Domain Intelligence

Is this domain two weeks old? Was it registered through a bulk registrar popular with cold emailers? Does its DNS configuration match known outreach infrastructure? These signals are invisible to a simple workflow but obvious to a purpose-built filter.

Continuous Learning

A workflow does the same thing every time. A real spam filter learns. It tracks which senders you engage with, which domains consistently send junk, and how spam tactics evolve. It gets better without you having to rebuild your workflows.

Studio Is a Productivity Tool, Not a Spam Filter

Google Workspace Studio is genuinely useful for what it's designed to do. Automating repetitive tasks across Workspace apps, connecting to third-party tools, building simple notification pipelines--these are real productivity wins.

But Google is marketing Studio's email templates as if they can solve your inbox problem. The "Notify me about urgent emails" workflow isn't email triage. It's a keyword matcher with a Gemini wrapper. And in 2026, when more than half of inbox spam is AI-generated, that's not even close to enough.

If you're evaluating Studio for email management, use it for what it's good at: automating workflows after your inbox is clean. Don't expect it to do the cleaning.

How Email Ferret Handles What Studio cannot

Email Ferret is built specifically for the problem Studio fails at: detecting and routing AI-generated cold outreach and spam before it clutters your inbox.

• Always-on monitoring: Every incoming email is scored in real-time--no workflows to build or maintain
• Multi-dimensional scoring: Dozens of weighted signals evaluated simultaneously, not one at a time
• Automation fingerprinting: Detects cold outreach platform headers and infrastructure patterns
• Domain intelligence: Evaluates sender domains for age, reputation, and outreach indicators
• Transparent decisions: See exactly why each email was classified, with detailed score breakdowns

Studio automates tasks. Email Ferret protects your inbox. They solve different problems--and your inbox needs both. To take immediate action on the cold outreach that slips through, see our guide on how to block cold emails in Gmail.

Frequently Asked Questions

Can Google Workspace Studio filter spam from Gmail?

No. Google Workspace Studio is a workflow automation tool, not a spam filter. It can build simple rules around email notifications and routing, but it lacks heuristic scoring, domain intelligence, and automation fingerprinting. In our test of 100 emails, Studio flagged 60% of cold outreach as "urgent" and missed more than half of genuinely important messages.

Is Google Workspace Studio better than Gmail's built-in spam filter?

They do different things. Gmail's built-in filter catches traditional spam (phishing, malware, known bad senders). Studio adds workflow automation on top of Gmail. Neither catches modern AI-generated cold outreach or BDR campaigns that pass authentication checks and use warmed domains. You need a dedicated detection layer for that.

Does Google Workspace Studio use Gemini to detect spam?

Studio uses Gemini to build and run workflows, but Gemini is a general-purpose language model--not a spam detection engine. It can classify text at a surface level, but it doesn't have access to domain reputation data, sender history, automation platform fingerprints, or the multi-signal heuristic scoring required to catch sophisticated cold outreach.

what is the best way to stop AI-generated cold outreach in Gmail?

Purpose-built email filtering tools that combine heuristic analysis, automation fingerprinting, and domain intelligence are the most effective. These tools monitor your inbox in real-time and catch the sophisticated cold outreach that Gmail's filters miss, AI assistants can't stop, and workflow tools like Studio get tricked by.

Can I use Google Workspace Studio and Email Ferret together?

Yes. They complement each other. Use Email Ferret to keep spam and cold outreach out of your inbox, then use Studio to automate workflows on the clean emails that remain--like summarizing threads, extracting action items, or routing messages to the right team channel.