Skip to main content

I Let AI Draft My Emails for a Week. Here's What It Actually Saved — and What It Ruined

A person typing on a laptop in a bright, modern office setting, showing productivity and technology.

Photo by cottonbro studio on Pexels

The pitch versus the reality

Every email client wants you to believe AI drafting is basically done. Gmail's got it, Outlook's got it, half the browser extensions in the Chrome store promise to write your replies in "your tone" before you've finished your coffee. I've written before about the typing shortcut hiding in your phone's keyboard, and this felt like the natural next test: if AI can finish a sentence, can it finish an entire email?

So I turned on AI drafting — Gmail's "Help me write" and a couple of the Outlook/Copilot suggestions — for every non-trivial email I sent over one work week. Not the one-word confirmations. The actual emails: scheduling, declining things, explaining a delay, following up on something I'd been ignoring.

The short version: it saved real time on maybe half of what I sent, and it actively made two emails worse in ways I had to catch before hitting send. That split is the interesting part, not the average.

Where it genuinely saved minutes

The AI draft tools are good at one very specific job: turning a two-line thought into a properly formatted, appropriately polite email when you already know exactly what you want to say. Think of the messages where the content isn't hard, just the phrasing is tedious.

Say you need to tell a vendor their invoice is late and you want them paid by Friday. You know that in one sentence. Typing the "hope you're doing well" and "please let me know if you have any questions" scaffolding around it is the actual friction. That's where AI drafting earns its keep — it fills in the social wrapper instantly, and you edit the middle.

Over the week, this covered:

  • Scheduling replies ("does Tuesday at 2 work, otherwise Thursday morning")
  • Short status updates to people who just needed a yes/no plus context
  • Declining a request I'd already mentally decided on

For these, I'd estimate I cut the actual writing time by more than half. Not because the AI is a better writer than me, but because it removes the blank-page problem for messages that don't need real thought — just phrasing.

Where it fell apart

The tools struggle the moment an email needs judgment instead of phrasing. Two examples from the week:

The delay explanation. I needed to tell a client a project was running late, and the honest reason involved a scheduling conflict on my end. The AI draft turned this into something vague and slightly evasive — lots of "unforeseen circumstances" language — because it doesn't know the actual reason and it defaults to safe, generic phrasing when the input is thin. It read like an email nobody would trust. I ended up writing that one from scratch.

The pushback email. Someone had asked for a scope change I needed to say no to, but nicely, with a specific counter-offer. The draft agreed with them. Not maliciously — it just optimized for "sounds agreeable and pleasant" because that's what most of its training examples of professional email probably look like. I caught it before sending, but if I'd been moving fast that morning, that's the kind of mistake that costs you a client relationship, not just an afternoon.

That second one is the one that actually worries me about this category of tool. It doesn't fail loudly. It fails by sounding confident and reasonable while quietly saying the wrong thing.

The pattern: length of thought, not length of email

By day three I had an informal rule that made the whole week make more sense: AI drafting works in proportion to how much of the email is "how do I say this" versus "what do I actually think." Short emails aren't automatically easy, and long emails aren't automatically hard — a five-line rejection with a specific counter-offer needed more of my brain than a ten-line project update where nothing was actually in question.

If you're just formatting a decision you've already made, let the AI dress it up. If you're still working out what you actually think, write it yourself, or at most use the AI for a first pass you plan to gut and rebuild.

A worked example of the actual workflow that stuck

Hands sketching in a notebook beside a laptop on a desk, capturing creativity.

Photo by Yan Krukau on Pexels

Here's roughly what I settled into by the end of the week, and what I've kept doing since:

1. Type a rough one-to-two sentence version of what I want to say, typos and all. 2. Let the AI draft expand it into a full email. 3. Read it once specifically checking: does this still say what I meant, or did it get smoothed into something softer or vaguer? 4. Edit the middle third — that's almost always where the actual content lives — and leave the opening/closing lines mostly as generated.

That's maybe 20-30% less typing on the emails where I already knew my answer, and roughly zero savings — sometimes negative, once you count the re-read — on the ones where I didn't.

Why this is different from a grammar checker

It's worth being clear about what's actually new here versus what's just repackaged. Grammar and phrasing tools have been catching typos and awkward sentences for a long time; that's a real but narrow job. Drafting tools are doing something else — generating the entire argument of the email from a prompt, including things you didn't explicitly ask for, like tone and level of agreement. That's a bigger surface area for it to get subtly wrong, which is exactly what happened with my pushback email. The convenience and the risk come from the same place.

FAQ

Does AI email drafting actually save time, or is it a wash once you factor in editing?

It depends entirely on whether you already know what you want to say. For routine replies where the content is settled and you're just fighting the blank page, it's a real time saver. For anything involving a decision, disagreement, or bad news, expect to spend nearly as long editing as you would have spent writing, and sometimes longer if you're not paying close attention to what it quietly changed.

Is it safe to use for anything sensitive, like HR or client conflict emails?

I'd be cautious. The failure mode I ran into — sounding confident while softening or misrepresenting what you actually meant — is exactly the kind of thing you don't want in a message where precision matters. Use it for structure and phrasing, but read every sentence of a sensitive email before sending, not just skim it.

Which is better, Gmail's built-in tool or a separate AI writing app?

For most people, the built-in option in whatever email client you already use is the practical choice — no copy-pasting between apps, and it already has some context about the thread. A separate app might write slightly better prose, but the friction of switching tools usually cancels out any quality gain for day-to-day email.

The boring conclusion, again

This is the same lesson I keep landing on with every AI feature I test: the tool is genuinely useful for the parts of a task that are mechanical, and genuinely risky for the parts that require judgment, and most people don't clearly separate those two things before they hit send. AI email drafting didn't replace my thinking, and honestly, I wouldn't want it to. It just took over the part of writing I never liked doing anyway — and that's a smaller, more useful win than the pitch makes it sound, but it's a real one.

Keep reading

Comments

Popular posts from this blog

Every To-Do App Works for the First Two Weeks — Here's What Decides the Rest

Photo by RDNE Stock project on Pexels The Honeymoon Phase Is Not the Test Here's a pattern I've watched play out dozens of times, including in my own phone: someone downloads a new task manager, spends a Saturday afternoon setting it up beautifully, and for about two weeks it's genuinely great. Everything's captured. Everything's organized. Then, sometime around week three, tasks start piling up unchecked, half the projects are stale, and the app quietly turns into another icon you feel guilty about. If this sounds familiar, your instinct is probably to blame the app. Todoist wasn't flexible enough. Things was too rigid. Notion took too much setup. So you switch, rebuild everything, and get another two good weeks. I've seen people cycle through four or five systems in a year this way, and each time they walk away more convinced that the "right" tool is still out there somewhere. It usually isn't the tool. The apps that survive long-...

The 20-Minute Weekly Reset That Keeps a Productivity System From Rotting

Photo by RDNE Stock project on Pexels The system isn't broken. It's just never reviewed. Here's a pattern I've watched play out with almost every productivity app I've tried: week one, it's magic. Everything gets captured, tagged, scheduled. Week three, there are 40 tasks with no due date sitting in an "Inbox" that nobody opens anymore. Week six, you've quietly gone back to a sticky note. I've written before about how every to-do app works great for the first two weeks and then something decides whether it survives past that. What I didn't spell out then is what that "something" usually is. It's rarely the app. It's the absence of a recurring moment where you actually look at everything you've captured and decide what to do with it. That moment has a name in productivity circles: the weekly review. It sounds like homework, and most explanations of it make it sound like a 90-minute audit of your entire l...

I Let an AI Draft My Email Replies for a Month. Here's What It Got Right (and Where I Still Type It Myself)

I used to treat "AI can write your emails now" as a solved problem, the kind of feature you turn on once and never think about again. Then I actually tried leaning on it for a full month — not just the odd smart-reply suggestion, but letting an AI draft actual responses to actual people — and it turned out to be way more uneven than the demos make it look. This isn't a "these five prompts will change your life" post. It's closer to a field report. Some of what I found will save you real time. Some of it will just annoy you, and I'd rather tell you that upfront than let you find out the hard way in a reply-all thread. Photo by cottonbro studio on Pexels What I actually tested I used the AI drafting features already built into Gmail and Outlook, plus a general-purpose assistant (Claude or ChatGPT, pasted in manually) for anything longer or trickier. No separate email-AI startup, no browser extension — just tools most people already have access to. ...