LinkedIn Headline Generator Prompt
Six headline options across three strategies, with character counts, plus the three lines of the About section that sit above the fold. The headline is 220 characters that follow you everywhere on LinkedIn, and almost all of them are spent on the same formula. This prompt writes six that are not: claim-led, proof-led and role-led, so you can pick the one that matches what you want the reader to do.
You write LinkedIn headlines. You have a strong opinion: a headline should make one specific claim a stranger can evaluate in four seconds, and most headlines make none.
## Me
{{ROLE}} at {{COMPANY}}
What we do: {{PRODUCT}}
Proof I can cite: {{PROOF}}
Who should stop and read: {{AUDIENCE}}
What I want them to do: {{OBJECTIVE}}
Tone: {{TONE}}
## Constraints
- 220 characters maximum, and print the count for every option.
- No "I help [X] do [Y] through [Z]".
- No pipe-separated noun lists ("Founder | Speaker | Investor | Dad").
- No emoji, no rocket, no "passionate about", no "on a mission to".
- Every claim must be supported by what I gave you above. If a strong headline would need a number I did not supply, write the version without it and tell me which number would improve it.
## Method
Write two options in each of three strategies:
- Claim-led: leads with the specific outcome we produce. Strongest when the outcome is unusual or measurable.
- Proof-led: leads with the credential or customer that makes a stranger believe the rest. Strongest when the category is crowded.
- Role-led: leads with what I do day to day, plainly. Strongest for job seekers and for people whose title is the thing being searched for.
Then write the About opening — the first three lines, which is all that appears above the "see more" fold. First person, no throat-clearing, no "as a seasoned professional".
## Output format
### Claim-led
Two options, each with (n characters).
### Proof-led
Two options, each with (n characters).
### Role-led
Two options, each with (n characters).
### Recommended
One, with one sentence on why given my objective.
### About — above the fold
Three lines. Then the count of characters in the first line, since that is the one that shows in most previews.
### The number that would make this sharper
The single missing fact.- You need a headline now and do not need a full audit.
- You have changed roles and the old one is stale.
- You are testing whether a sharper claim changes your acceptance rate.
- Fill {{PROOF}} even if the answer is "none yet" — it decides whether the proof-led options are honest or invented.
- Pick by objective, not by taste. The claim-led headline wins for outbound; the role-led one wins for job search.
- Check the character count manually before saving. Models miscount, and LinkedIn silently truncates.
The output schema it demands
The single highest-leverage part of any prompt. Without a declared shape you get prose you then have to reformat; with one, you get sections you can act on — and you can tell at a glance when the model skipped something.
### Claim-led (2 + counts) ### Proof-led (2 + counts) ### Role-led (2 + counts) ### Recommended ### About — above the fold ### The number that would make this sharper
What it looks like filled in
Illustrative, not a promise — this is the shape of a good answer rather than a guarantee of one. The scenario is the same fictional company used across the whole toolkit.
Founder at Cadence. Objective: a prospect who gets a connection request decides I am worth replying to. Proof: two logistics customers, one named.
Claim-led — "Founder at Cadence — we cut multi-entity payroll close from 9 days to 2" (68 characters).
Proof-led — "Founder at Cadence. Built for the finance teams other software calls an edge case: 3+ entities." (98 characters)
Recommended — claim-led. For outbound specifically, a concrete number does more work in four seconds than a credential does.
The number that would make this sharper — how many entities the named customer runs. "9 days to 2" reads stronger next to "across 5 entities" than alone.
The reference behind this prompt
The two Premium tiers differ mostly in who can see you and how far back. Relevant if the headline change is part of a job search.
Written to a fixed structure and run through Claude and ChatGPT to confirm each one returns the output format it declares. That is not the same as a measured reply-rate benchmark, and nothing here claims one — if you run these at volume and have numbers, we want them.
Take the whole folder instead
This prompt, the files it expects and the rest of the sequence, as a zip you can drop into a Claude Project, a Cursor workspace or an agent repo. Free, no account.