You have a colleague, or a few, who has clearly figured something out about email and AI.
They reply faster than they used to. They mention "letting the AI take a first pass" in the same casual tone you might mention pouring coffee. Their inbox does not seem to take the hours out of their week that yours does. You do not see the seams.
You, meanwhile, have tried ChatGPT for an email maybe twice. The first time it sounded like a press release. The second time you spent more time writing the prompt than you would have spent writing the email. You closed the tab and went back to typing.
The gap feels real. It is also probably smaller than you think.
What "behind" actually costs
If you send 20-30 substantive emails per workday (the kind that take more than 30 seconds to compose), and you are typing every one of them from scratch, you are spending somewhere between 60 and 100 minutes a day on the literal mechanics of email composition. Not on the thinking. On the typing, the re-reading, the hunt for a slightly better word, the second-pass proof.
A reasonable AI workflow takes 10-20 minutes off the top of that. Not by replacing your judgment. By absorbing the part of the task that does not need judgment: the structure, the polite scaffolding, the first competent draft of "yes, Tuesday at 3pm works, I'll bring the deck."
Over a five-day week, that is something between 50 minutes and three hours. Compounded across a year, the gap between using AI for email and not using it is real. It is also addressable in an afternoon.
Why the gap looks bigger than it is
The people who have figured out AI for email almost universally got there by way of a long, frustrating road: ChatGPT system prompts that took twelve drafts to get right, copy-paste workflows that required tab-switching for every reply, a custom GPT they tried to train on their own writing that never quite captured their voice. They are happy with where they ended up. The road is not what they would recommend.
If you are catching up now, you can skip the road. The category has matured. You do not need to become a prompt engineer. You do not need to build a workflow that involves three browser tabs and a pasted system message. The good tools have absorbed the complexity into the product.
Two things have changed in the last year that compress the catch-up workflow:
The good tools live in your inbox now. Instead of opening a separate chat window, pasting an email, writing a brief, copying the result, and pasting it back, the modern shape is a draft button inside Gmail or Outlook that reads the thread automatically and produces a draft in place. The cognitive load is closer to "click, skim, send" than "compose a brief for a junior assistant." This is the in-inbox workflow we covered in Stop Tab-Switching: The In-Inbox AI Workflow That Beats ChatGPT for Email.
The output has caught up to the point where it is worth using. The honest critique of AI email two years ago was that the drafts sounded like a corporate brochure. They still do, in tools that have not invested in voice matching. But the tools that read your sent email history and learn your patterns produce drafts that read recognizably like you, on most everyday email, after about a week of background data collection. The quality threshold for "I would actually send this" is finally clearable, not just claimed.
The catch-up workflow, in one afternoon
This is the version of the setup that ignores everything you have heard about prompt engineering. You do not need any of it.
Step 1: Install one in-inbox AI tool. Pick one that runs inside Gmail or Outlook (whichever you use) and one that has a meaningful free tier so you can test it on real email without committing. The whole thing should take five minutes from "click install" to "first draft in your reply box."
Step 2: Use it on five real emails today. Not test emails. Not contrived examples. Real replies you owe people. Click the draft button, read what it produces, edit if needed, send. The point is not to evaluate the tool yet. The point is to feel the workflow.
Step 3: Notice which drafts you sent without changing anything. This is the metric that matters, the one we covered in Every AI Email Tool Saves Time. The Real Question Is Whether You'd Send It Without Reading.. On day one, with a tool that has not yet seen any of your sent email, the rate will probably be low. After a week of background data collection by a tool that does voice matching, it should climb meaningfully.
Step 4: Adjust your expectations to match the curve. AI email is not a one-shot upgrade. It is a tool that gets better with two things: data on how you write (which a voice-matching tool collects in the background) and your own calibration on which kinds of emails it is good at versus which ones still need you to write from scratch. Both happen naturally over a week or two of normal use. You do not have to do anything special.
That is it. There is no prompt to memorize. No system message to maintain. No workflow that requires rebuilding your relationship with your inbox.
What you do not need to do
You do not need to watch a course on prompt engineering. The good in-inbox tools handle the prompting for you, conditioned on your sent email history.
You do not need to switch email clients. Tools that work inside Gmail or Outlook as extensions are now better than tools that require you to migrate to a new email app, for most use cases.
You do not need to become a "ChatGPT power user" first. The whole reason this category has matured is so you can skip the ChatGPT phase entirely and get to the working workflow without paying the prompt-engineering tax.
You do not need to commit to a paid plan to find out if it works. The honest in-inbox tools have free tiers that are generous enough to test against your real inbox for a week.
The narrow window where the gap actually matters
The compounding effect of this is not that you save an hour a day starting tomorrow. It is that the people who close the gap stop carrying email as a cognitive load. They stop dreading the inbox. They stop using their highest-attention hours for typing replies that an AI could have drafted in their voice in two seconds.
You do not need to be ahead of anyone. You need to be in the workflow, getting the compounding benefits, instead of outside it watching other people get them. The catch-up is shorter than the people inside it remember it being.
ForthWrite runs inside Gmail and Outlook, reads your sent email history (encrypted, isolated to your account), and starts producing voice-matched drafts on day one. The first 10 drafts per week are free, no API key required, no prompt-engineering experience needed.
If you want to start with a no-install option that lets you feel the voice-matching idea before committing to a tool, the Persona Prompt Generator builds a first-person prompt from your answers in about five minutes. Free, no account required.