Which tool to choose
For a follow-up email you don't need a paid tool: the task is short and doesn't require advanced reasoning.
- ChatGPT free: the right choice in 99% of cases. Free accounts can search for up-to-date information on the web, analyze data, upload images or files, and use GPTs: a solid kit for light personal use. The real limit: with the most powerful model you can send a limited number of messages in a window of a few hours, then the chat falls back to a lighter version until the reset. For two or three emails a day you never touch it.
- Gemini or Claude (free versions): equivalent for this task. If you already use one of the two, stay there: switching gives you no advantage on such a short text.
- Plus at $20/month: useless for this. It's needed if you generate dozens of emails a day and run into the limits, or if you use it for completely different work.
Choose the one you already have open. For the follow-up they're interchangeable.
How to do it
From browser or app, the path doesn't change:
- Open the tool and start a new chat.
- First gather three pieces of information the AI can't invent: the exact name of the contact, the role you applied for (or the subject of the meeting), and a concrete detail that came up during the conversation (a project mentioned, a team challenge, a specific question).
- Paste the prompt below, filling the parts in quotes with your data.
- Read the draft. Check that the concrete detail is yours and not a generic filler.
- Correct names, dates, and any number. The AI sometimes rounds or invents: this part you verify yourself, always.
- Paste it into your mail client, reread it aloud once, send.
The operational syntax:
Write a follow-up email after a job interview.
Language: Italian. Tone: professional but warm, never servile.
Maximum length: 130 words.
Data:
- My name: "Marco Bianchi"
- Person I'm writing to: "Dr. Laura Rossi, recruitment manager"
- Role I applied for: "Customer Success Specialist"
- Interview date: "June 9"
- Real detail that came up in the interview to recall: "the project to
reduce churn on enterprise clients that she told me about"
Structure: thanks for the time, recall of the real detail,
a sentence restating my interest, availability to provide
more information, closing. Email subject included, short and
specific with name and role.
If the draft exceeds 150 words or sounds flat, reply in the same chat:
Cut it to 100 words and make the reference to the project more specific.
It rewrites it for you right away.
On send timing: the thank-you should be sent early. Send the first email within one or two days of the interview, ideally the next morning, so they see it at the start of the day; avoid the same day, because the recruiter may be busy with other candidates. If instead you're chasing a reply that isn't coming, the typical wait is 5-7 working days when you weren't given a timeframe; if instead they indicated a date for the decision, wait until it has passed and write 1-2 working days after.
Concrete example
Giulia had an interview on Tuesday for a copywriter role. During the meeting the manager mentioned that the team was about to launch a website rebranding. Wednesday morning Giulia opens ChatGPT for free and pastes the prompt, putting "upcoming website rebranding" as the real detail.
The draft comes out at 100 words, with the subject "Giulia Conti – Follow-up Copywriter interview of 6/9". One problem: the AI had written "your imminent September rebranding", inventing the month. Giulia deletes "September" and leaves only "the website rebranding we talked about". She also changes a sentence that's too enthusiastic ("I would be honored") to "I would be glad to contribute". She rereads, sends at 9:15.
Result: the manager's reply that day, with confirmation of the second interview. Total time spent by Giulia: seven minutes.
When it does NOT work (and how to fix it)
If the AI invents details of the interview
The model fills the gaps with things that are plausible but false: a project never mentioned, a nonexistent product name. Fix: provide the real detail yourself in the prompt (as the template does) and never ask the AI to "imagine what you discussed". If you don't have a specific detail, add to the prompt:
No specific detail to recall: stay generic, without
inventing projects or names.
You'll get a more neutral email, but an honest one.
If the email sounds fake or fawning
Phrases like "your cutting-edge company" give the AI away from a distance. Reply in the chat:
Remove every promotional adjective and praise of the company. Write
the way a real person would speak, dry.
It works almost always on the first try.
If it's too long
A recruiter reads diagonally. Long emails are a mistake: keep it under 150 words and respect their time. Impose the length in the prompt and, if needed, ask for a further trim.
If you've already written twice without a reply
Here the problem isn't the email, it's the insistence. Send a single follow-up, unless they expressly invited you to continue the exchange. A final note after five to seven days, then you stop: continuing beyond that works against you.
A tip from someone who really uses it
The AI gives you the structure in ten seconds, but the value of the email lies in the one thing it doesn't possess: the specific detail of what you said to each other. When two candidates are equal and the recruiter can't decide, a follow-up with a genuine recall can tip the scales. So as soon as you leave the interview, before even opening ChatGPT, write on a note two things the contact said that struck you. Those two lines, slipped into the prompt, turn an anonymous email into one the recruiter remembers.
Frequently asked questions
How long should the follow-up email be?
Short. One paragraph, two at most. Aim for 100-130 words. Beyond 150 you risk it not being read to the end.
Should I write to everyone who interviewed me?
Yes, if there was more than one. Send an email to each: the same day is ideal, the next morning is fine. Personalize each one with a different detail that came up with that specific person, otherwise they notice it's a copy-paste.
Can I use the same email even if they rejected me?
It's worth replying anyway. A gracious reply after a rejection is rare and keeps the door open for future roles or referrals. Change the prompt indicating:
Reply graciously to a rejection: thank them for the time and ask
to stay in touch for future opportunities.
Will the recruiter realize I wrote it with AI?
They realize it only if you leave it as it came out: too polished, full of superlatives, with brochure-like enthusiasm. The point isn't to hide the tool, it's that the email sounds like you. That's why the steps above insist on removing the inflated praise and on inserting a detail no model could know. A dry, specific email sent at the right moment doesn't have "the smell of AI": it has the smell of a person who cares.