How to Write a Follow-Up After a Networking Chat Without Cringe
An 8-minute writing pass for a 5–7 line networking follow-up that names one real moment and one concrete ask. Ban LinkedIn-bro fluff. ChatGPT or Claude.
Anyone who just had a real networking chat can write a follow-up email without cringe in about 8 minutes: name one specific moment from the conversation, add one concrete next ask, and keep the note to five to seven lines. Run it in ChatGPT or Claude before the model pads the message with LinkedIn-bro fluff.
A networking follow-up email is a short receipt of a real moment plus a clear next step, not a personal brand essay and not a “circle back” template. If the model invents a compliment you never earned in the chat, you never did the job. One specific moment plus one ask beats a paragraph of soft networking fog.
This is not an apology note that sounds like you. An apology repairs harm. A follow-up continues a useful thread. It is not a calendar invite people actually open. A calendar invite schedules time. This email earns the right to ask for time. It is not a customer quote into a one-sentence testimonial. Testimonials are public praise. Follow-ups are private next steps.
What you need
Your rough notes from the chat: one specific moment, one shared topic, and one next ask you could actually honor. ChatGPT or Claude. Eight minutes and one cold read out loud as if you are the recipient.
Banned before you start: “great connecting,” “synergy,” “would love to pick your brain,” fake intimacy, invented compliments, multi-ask laundry lists, and LinkedIn-bro fluff that could fit anyone.
Write the five to seven line follow-up
1. Paste only the real moment and the real ask
Paste three bullets max: who they are in one noun phrase, the specific moment (a story, tool, or opinion they shared), and the one next ask (a 15-minute call, a link, an intro you already cleared).
Instruct ChatGPT or Claude: “Write a 5–7 line networking follow-up email. Name one specific moment from my notes. Include exactly one concrete next ask. Ban LinkedIn-bro fluff. Ban inventing compliments not in my notes. Ban ‘pick your brain.’ Then stop.”
Good notes: “She mentioned the warehouse inventory pilot last Tuesday; ask for the one-pager she offered.” Bad: “She seemed inspiring and visionary.” Vague praise is how cringe starts.
2. Force five to seven lines with one ask
Shape the email as: greeting, one-line thanks tied to the moment, one-line proof you listened, one concrete ask with a time box, optional one-line offer, short sign-off. Cap at seven lines of body.
Instruct: “Output exactly 5–7 short lines in the body. One ask only. No buzzwords. No invented details. Keep my voice plain.”
Good: “Thanks for walking me through the inventory pilot — the barcode exception case stuck with me. If the one-pager is still handy, could you share it this week?” Bad: “Loved your energy and would cherish the chance to deepen our synergy.” Synergy is a hard fail.
Keep the moment checkable
If a friend who was not in the chat cannot tell what happened from your sentence, rewrite it. Specificity is the anti-cringe filter.
3. Ban LinkedIn-bro fluff and invented intimacy
Tell the model: “Delete any line that could be pasted to a stranger. Delete ‘excited,’ ‘thrilled,’ ‘circle back,’ ‘thought leader,’ and fake shared history. Do not invent that we laughed if my notes do not say so.”
ChatGPT will add “it was wonderful connecting.” Claude will soften into “I so appreciated your insights.” Both are fog. Cut them. Replace with the moment.
Hard fail: brain-picking, multi-asks, invented jokes, status flattery, and any line longer than two short sentences that says nothing concrete.
4. Cold recipient read before you send
Read the draft out loud as if you are busy and did not ask for this email. If you would skim-delete it, shorten. If the ask is vague, make it binary or time-boxed.
Check that the specific moment appears once and the ask appears once. Two asks is how follow-ups die. Move the second ask to a later email after they reply.
5. One human pass, then send or stop
Send only if the moment is true and the ask is something you would honor if they say yes. Do not let the model “warm it up” with fluff after your cut.
Check traps: fluff phrases, invented compliments, missing moment, missing ask, more than seven lines. Fix with: “Rewrite only the lines I mark. Keep 5–7 lines. One ask. No fluff.”
One pass. If the note still sounds like a template for anyone, you never wrote a follow-up after a networking chat without cringe. Stop.
What good looks like
Trigger: You are about to send “great connecting, would love to pick your brain sometime.”
Input: Real moment notes, one ask, ChatGPT or Claude, cold recipient read.
Output: 5–7 line email naming one specific moment and one concrete next ask, zero LinkedIn-bro fluff, zero invented intimacy.
Stop: After one cold read. No second ask. No brand essay.
Treat the email like a receipt, not a performance. Busy people reward proof of listening. Models reward polish. Prefer proof. Keep names and project titles exactly as they said them. Correcting their product name in your follow-up is how you look like you were not there. When models “helpfully” add mutual connections you never mentioned, delete them. The cold recipient read exists to catch fluff and missing asks. Skip that read and you will send fog. That wastes the chat.
Takeaways
- Paste one real moment and one real ask. Force 5–7 lines only.
- Ban LinkedIn-bro fluff. Ban inventing compliments not in your notes.
- Keep the moment checkable by a stranger reading the email.
- One cold recipient read. Then send or stop.
Keep the subject line boring and specific: reply about the inventory one-pager beats circling back on our chat. Boring subject lines get opened. Clever ones get skimmed. Models love clever. Prefer boring.
Frequently Asked Questions
How soon should I send the follow-up?
Same day or next morning while the moment is fresh. Waiting a week forces you to invent details the model will gladly fake.
Can I ask for a coffee and an intro in the same email?
No. One concrete ask only. A second ask waits until they reply to the first.
Should I use ChatGPT or Claude?
Either works if you ban fluff and inventing. Claude is steadier at plain tone. ChatGPT needs a hard ban on “thrilled to connect.”
What if I forgot to take notes during the chat?
Write three bullets from memory before opening the model. If you cannot name one specific moment, wait until you remember — do not let the model invent one.
Is a LinkedIn comment enough instead of email?
A public comment is not this job. This guide is for a private 5–7 line follow-up with one ask.