Walk me through your current AI-for-email workflow.
You are in Gmail. A thread lands that needs a real reply. You open a new tab, navigate to chat.openai.com. You paste the original email. You write a brief. You copy the output. You switch back to Gmail. You paste it. You read it, realize it is a bit off, switch back, adjust the prompt, copy again.
From email to sent: maybe 3-5 minutes. For one email.
If you send 20-30 emails that need this treatment per day, you are spending 60-150 minutes a day in this workflow. Not drafting. Switching. And switching is cognitive overhead, not just time.
What the tab-switching workflow actually costs
Time cost
The obvious cost is the literal minutes. 3-5 minutes per email, 20 emails, 5 days a week. That is 5-8 hours per week spent on the mechanics of the workflow, not the thinking.
Attention cost
Context-switching between a ChatGPT tab and Gmail fragments the attention that makes email replies good. When you are drafting a sensitive reply to a client who gave ambiguous feedback, the last thing you want to do is leave the thread, write a brief to an AI, come back, and then have to re-establish context. The good reply often requires staying in the thread, feeling the emotional register of it, and responding from inside that context.
Compounding distraction
Every time you open a new tab, you are one click away from a distraction. The notification badge on the browser. The LinkedIn tab already open. The email from someone else you saw while switching back to Gmail. The tab-switching workflow compounds distraction into your email workflow.
What in-inbox AI looks like
An in-inbox AI email workflow does not require you to leave the email. It sits inside Gmail or Outlook as an extension, reads the thread you are looking at, and generates a draft when you ask for one, in the same window, with zero context-switching.
The best in-inbox tools do three more things that tab-switching ChatGPT cannot:
1. They know the thread. The AI reads the full email thread automatically. You do not have to paste anything. The relationship history, the last thing the person said, the topic: all available without a brief.
2. They know your voice. A trained in-inbox tool has seen your sent email history. Its outputs carry your specific patterns (openers, closers, length, tone) because it learned them from your actual sent folder, not from a system prompt you wrote.
3. They filter noise automatically. A good in-inbox AI does not try to draft a reply to your Amazon shipment notification. It recognizes transactional, promotional, and newsletter email and skips them. You only interact with the AI when it is useful.
The 12-second version
The fastest version of this workflow: open an email thread, click the FinalDraft button in the compose panel, see a draft appear in 2-3 seconds that is conditioned on the full thread and your sent email history, make a small edit if needed, send.
12-15 seconds from "open thread" to "draft ready." Compare that to the tab-switching workflow.
The quality is not the same as a carefully crafted ChatGPT brief; the brief lets you give the AI more specific direction. But for 70-80% of your email volume, the tab-switching workflow is over-engineering the task. You do not need to write a brief for "reply to this person who is following up on a status I owe them." You need a competent, in-voice draft that you can skim, possibly edit the last sentence, and send.
When tab-switching is still worth it
The in-inbox workflow is not always better. For genuinely high-stakes emails (board communications, difficult client conversations, new pitches that need to be perfect), the ability to iterate with ChatGPT over multiple rounds, with manual direction, is still valuable. Use the right tool for the moment.
But the 80% of your email volume that is status updates, scheduling, follow-ups, acknowledgments, and straightforward requests does not need that treatment. It needs a good draft, fast, that sounds like you wrote it.
FinalDraft runs inside Gmail and Outlook. No tab-switching required. The first 10 drafts per week are free, no API key needed. Try it with an email that is sitting in your inbox right now.
Or start with the Persona Prompt Generator if you want to build a voice-matched prompt you can use in ChatGPT while you evaluate.