A while back I posted that we know so much about every target decision-maker it is almost scary. That line was not a brag. It was the honest description of what my outbound looks like now that an AI email writer sits in the middle of it. I run Growth Cab, a GTM advisory, and I send outreach every week for real accounts with real revenue on the line. So this is the version nobody selling you a tool will give you: where an AI email writer actually earns its place, and where it quietly makes you worse.
in
“We know so much about every single target decision-maker that it's almost scary. One of the most powerful workflows we run at Growth Cab combines fresh contact data with Claude Code to automate highly personalized outreach.”
Most of What an AI Email Writer Produces Is Faster Spam
Here is the uncomfortable part. The default way people use an AI email writer is to take a shaky template, ask the model to spin fifty variations, and blast them. The output reads fine. It is grammatical, confident, on brand. It is also generic, and your prospect has already read the same email nine times this month from nine other tools running the same trick. The model did its job. The job was just the wrong one. You did not buy leverage. You bought the ability to be ignored at a higher volume.
The reason is simple. A message is only as specific as the information behind it. If you feed an AI email writer nothing but a name, a title, and a company, it will write to a name, a title, and a company. It fills the gaps with fluent filler because filler is all you gave it. Volume was never the constraint in outbound. Relevance was. An AI email writer that only scales volume scales the one thing that was never broken.
The Test I Use for Any AI Email Writer
I have one test and it decides everything. Take a message the tool wrote and imagine sending the exact same words to a different person at a different company. If it still makes sense, it is filler and it goes in the trash. If it would make no sense at all, because it references something true about this one account and no other, you have something worth sending. The goal is a message so specific it is non-transferable. That single rule kills most of what these tools produce by default, and it is the fastest quality check I know.
The Research Layer Is the Whole Game
The AI email writers that actually convert are not better at writing. They are better fed. Underneath the good ones there is a research step that runs before a single word gets drafted, and that step is where the value lives.
Ours starts with clean data. We pull from Prospeo, which gives us phone numbers and triple verified contact details for decision makers, refreshed inside the last week. Reaching the right person is a boring problem that quietly sinks most outbound, and solving it first means the writing is aimed at a human who actually exists and actually holds the budget. Prospeo is a paid tool we use and partner with, so read that as a disclosure rather than a neutral tip. The principle holds with any verified source: bad data makes even a great AI email writer polite and useless.
Then we let Claude do real research on each prospect before it writes. It reads their recent posts, their company, the shape of their role, and it hunts for one true reason this conversation is relevant to them this week. Only after that does it draft. The email that comes out does not open with a pitch. It opens with their situation, asks one real question, and leaves the calendar link in a drawer. The AI email writer is the last step in that chain and the least important one. The research is what makes the writing land.
How I Actually Run It
The setup is a division of labor. The AI email writer and the research stack handle the parts that are mechanical and repeatable: finding the contact, verifying it, reading the public trail, drafting a first version that belongs to one account. I handle the parts that are judgment: which accounts are even worth the effort, whether the angle is honest, and every reply once a human answers. The model brings coverage. I bring the decision about where to point it.
The numbers that matter to me are small on purpose. I do not measure how many emails went out. I measure how many started a real conversation and how many of those became a call. When the scoreboard is conversations, you stop admiring send volume and start caring whether each message deserved to exist. An AI email writer helps that goal only when it is aimed at a short, well chosen list. Aim it at a giant list and it becomes a very efficient way to burn your domain.
Where an AI Email Writer Still Breaks
I owe you the limits, because the demos will never mention them. First, thin inputs. If your data is weak or your positioning is vague, the tool inherits both and hides them behind fluent sentences. It does not fix a bad offer. It makes a bad offer readable, which is worse, because now it looks like effort.
Second, judgment about silence. An AI email writer is good at deciding what to say and terrible at deciding whether to say anything at all. It will happily write a confident, well researched message to an account any experienced rep would skip. Left alone it optimizes for activity, and activity against the wrong list is how you lose a domain and a reputation on the same afternoon.
Third, the conversation itself. The tool can open a thread. It cannot hold the human exchange that turns a reply into a booked call, read the hesitation behind a short answer, or know when to drop the script and just be useful. The moment a real person replies with something unexpected, a human has to take the wheel or the whole thing collapses back into the noise you were trying to escape.
So the honest verdict is a careful yes. An AI email writer is worth building around, as long as you treat the writing as the easy last step and spend your real effort on the data and the research in front of it. Get that order right and the same list you already have starts producing conversations. Get it backwards and you have paid for a machine that helps the wrong people ignore you faster.
If you want one play like this every morning, I write The Revenue AI Brief, a short daily note on the AI and GTM systems I run inside Growth Cab, including the prompts behind our outreach. Drop your work email below and tomorrow's edition lands in your inbox. And if you think I am wrong about where the AI email writer breaks, tell me on LinkedIn. I answer everything.
Disclosure: this article mentions Prospeo, a tool Growth Cab uses and partners with. The workflow and the opinions here are my own.
