A short LinkedIn message that gets a reply — specific ask, no "pick your brain".
プロンプト 入力欄 5 件
Write a LinkedIn DM to , who is a at . I want to eventually apply for there.
Constraints:
- Under 80 words
- Open with something specific about THEIR work:
- One concrete, answerable ask (a 15-min call OR one specific question — not both, never "any advice")
- Name why them specifically, not "someone at "
- No flattery words (amazing, incredible, huge fan), no apology for reaching out
- End so replying takes under a minute
Also write a follow-up for 7 days later if there's no reply (under 40 words, adds one new reason, no guilt).
出力例
DM:
Hi Maya — your post on cutting Figma handoff docs to a single annotated frame changed how our student design club ships. I'm targeting product design internships for summer and Linear is top of my list. One question: what did your portfolio show that you think actually got you the interview there? Happy with a one-line answer.
Follow-up (day 7):
Hi Maya — adding one thing since I wrote: I shipped the annotated-frame approach in our club's last project (link). Still curious about the portfolio question if you get 30 seconds.
ヒントThe specific-question ask outperforms the call ask roughly 3:1 for cold messages — calls cost 30 minutes, one-line answers cost 30 seconds. Ask for the call only after they have replied once.
ChatGPTGeminiClaude
Reply to a lowball offer (negotiate in writing)
Counter an offer by email: enthusiastic, numeric, and impossible to read as an ultimatum.
プロンプト 入力欄 6 件
Help me counter a job offer by email.
Offer: at , base , [OTHER COMPONENTS — bonus/equity/benefits].
My target: . My leverage: [MARKET DATA / COMPETING OFFER / SPECIFIC SKILLS — whatever is true].
Write the reply email:
- Open with genuine enthusiasm for the role (1 sentence, specific to the team/work, not "thrilled and honored")
- State the counter as a number, anchored to , not to my personal needs
- Ask about ONE non-salary lever too:
- Keep the door explicitly open — no "I need X or I walk" energy
- Under 150 words
Then add: the one sentence I should NOT include, with a one-line why.
出力例
Subject: Re: Offer — genuinely excited about this
Hi Dana,
Thank you — after meeting the analytics team I'm confident this is where I'd do my best work, and I want to make this happen.
On compensation: based on current market data for senior analysts in Austin (Levels and two recent competing conversations put the band at $128–140k), I'd be ready to sign at $134k base. If base has a hard ceiling, a signing bonus or a 6-month review with a defined path to that number would also work for me.
Everything else in the offer looks great. What flexibility do you have?
Best,
Sam
Do NOT include: "I was hoping for more because of my rent/loans" — personal-need anchors invite sympathy, not budget; market anchors invite counter-offers.
ヒントCountering in writing beats the phone for most people: you keep the number precise, nobody reads your nervous pause as flexibility, and the recruiter can forward your exact words (and your market data) to the approver.
ChatGPTGeminiClaude
Turn duty bullets into result bullets
Rewrite flat resume lines into quantified impact — and interrogate you for missing numbers.
プロンプト 入力欄 3 件
Rewrite my resume bullets from duties into results for a application.
Rules:
- Formula: strong verb + what changed + measured how much + for whom/at what scale
- If a bullet has no number, DO NOT invent one — instead ask me up to 2 questions per bullet that would surface a real number (how many, how often, compared to what, how big was the team/dataset/budget)
- Kill weak verbs: responsible for, helped with, worked on, assisted
- Max 24 words per bullet
- Keep terms from this job posting where honest:
My bullets:
出力例
1. "Responsible for the club newsletter"
→ Grew biweekly newsletter from 120 to 480 subscribers in two semesters; 42% average open rate (campus org average: ~25%).
2. "Helped with onboarding new members"
→ Need a number first: (a) How many members joined during your time? (b) Did time-to-first-contribution change after your onboarding doc?
3. "Worked on Python scripts for data cleaning"
→ Automated cleaning for a 40k-row survey dataset in pandas, cutting weekly prep from ~6 hours to 45 minutes for a 3-person research team.
ヒントThe interrogation half is the real value — answer its questions in a doc once, and you have a fact bank to quantify every future application honestly instead of guessing on deadline night.