Verified against ChatGPT · 2026-08-08
Turn a thin connection into a networking message that actually gets a reply
Drafts a short LinkedIn or email networking message tailored to exactly how you know (or don't yet know) the person, so it reads as a real ask instead of a mass-sent template.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are drafting a networking message I will send to one specific person — not a mass template, a message this one recipient will read and decide whether to answer within about ten seconds.
WHO I'M MESSAGING
Senior PM at a Series C fintech I'm targeting; two levels above the role I'd apply for.
HOW WE'RE CONNECTED
We both attended the same three-day product conference last month; no direct interaction there.
WHAT I ACTUALLY WANT
A 15-minute call about how their team structures the PM interview loop, not a job referral yet.
WHAT I CAN OFFER OR REFERENCE
I built a similar usage-based pricing model at my current company and can compare notes.
PLATFORM AND LENGTH
LinkedIn connection-request note, must fit under 300 characters.
RULES
Open with the specific, true reason I'm messaging this person and not a hundred others — reference the connection basis in the first sentence, not buried after a generic greeting. State the ask in one sentence, and make it a low-effort, specific ask (a 15-minute call on a named topic, an answer to one named question, a referral to a named role) rather than an open-ended "pick your brain" request that forces them to do the work of figuring out what I want. If the connection basis is weak or nonexistent (a cold LinkedIn message to a stranger), say so plainly in your reasoning and write the message to earn attention on the strength of the ask and the shared relevance alone — do not fabricate a warmer connection than actually exists. Never use flattery that could apply to anyone ("I've been following your incredible work") unless I gave you something specific to point to. Keep the message short enough to read in the platform's preview pane without a "see more" click — that means genuinely brief for LinkedIn, slightly longer is fine for email. End with a specific, easy next step the recipient can say yes to without scheduling overhead (a yes/no question, a suggested 15-minute window, a single link).
WHAT NOT TO DO
Do not open with "I hope this message finds you well" or any other line that signals mass-template. Do not ask them to "jump on a call to learn more about my background" — that puts the burden of structuring the conversation on them. Do not thank them in advance for something they haven't agreed to yet.
OUTPUT FORMAT
1. The message itself, ready to send, sized to the stated platform.
2. One alternate subject line or opening line if this is an email, so I can A/B the open.
3. A one-line note on what would make this message stronger if I have more specific shared context to add.Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
The instruction to open with the connection basis in sentence one exploits how people actually triage LinkedIn and email inboxes: they decide relevance from the first line before reading the ask, so burying the "why you" after a greeting means the message gets judged as generic before the real content is even seen. Forcing the model to name whether the connection is genuinely warm or actually cold prevents GPT-5.1's default politeness bias from inventing a rapport that doesn't exist — left unconstrained, it tends to write "great to reconnect" language for a recipient the user has never actually spoken to, which reads as either confused or dishonest the moment the recipient checks. Specifying that the ask must be low-effort and singular addresses a real failure mode of networking messages: an open-ended "would love to pick your brain" ask requires the recipient to invent the agenda, and busy people route effort-shifting requests straight to ignored rather than doing that work for a near-stranger. The platform-and-length constraint matters mechanically because LinkedIn truncates messages in the notification preview and connection-request notes have a hard character cap — a message written without that constraint in mind often gets cut off mid-sentence exactly where the ask was going to land. Ending on a single yes/no question or proposed time window rather than an open "let me know if you're free" closes the loop the way a specific calendar invite does compared to a vague "we should catch up sometime" — it converts a reply from an effortful scheduling negotiation into a one-tap response, which is the single biggest lever on reply rate for a message this short.
What you get back
Hi Priya — we were both at the SaaStr product track last month (I was on the pricing-model panel audience). I'm restructuring my company's usage-based pricing and would love 15 minutes to hear how your team handles PM interview loops for pricing-focused roles — not asking about openings yet, just comparing notes. Would a quick call next Tuesday or Wednesday work?
Verified against
ChatGPT GPT-5.1 · 2026-08-08
Changelog
- 2026-08-08 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
