Handle a Candidate's Counter-Offer Ask
When to use it
Prepare talking points before a call where a candidate is asking for more than the offered range.
The prompt
I extended an offer for a {{Job Title}} role and the candidate has come back asking for {{what they're asking for — higher base, more equity, a signing bonus, a different title, an earlier review}}. ORIGINAL OFFER: {{base, bonus, equity, sign-on, level}} THE APPROVED BAND AND WHERE THE OFFER SITS IN IT: {{e.g. 70th percentile of the band}} WHAT I ACTUALLY HAVE ROOM TO MOVE ON: {{be honest — budget, level, non-cash items, or nothing}} WHAT THEY SAID: {{their stated reasoning, any competing offer and whether I've seen it, current comp if they shared it}} HOW STRONG A CANDIDATE, AND HOW HARD TO REPLACE: {{your honest view}} INTERNAL EQUITY: {{roughly what people already on the team at this level earn}} Give me: 1. 2-3 talking points for the call that acknowledge their ask without committing to anything I haven't cleared. 2. Two questions that tell me whether this is their real bar or an opening position, and what actually matters most to them. 3. The non-salary levers I could trade if base is fixed — sign-on, start date, title, a written early review, remote flexibility, learning budget — and which suit this ask. 4. Any risk the ask creates: going over band, paying a new hire more than the team they're joining, or setting a precedent. 5. If I truly have no room, how to say so in a way that keeps the relationship and doesn't sound scripted. 6. A one-paragraph note to the hiring manager or comp team requesting approval, in case I need it. Don't invent market data. If you mention market rates, label them as a guess I need to verify.
Tip
Never say a number out loud that you haven't cleared internally. A promise walked back does more damage than a slow answer, so use this for the shape of the conversation and tell the candidate plainly when you'll come back to them.
How to use it
Paste the prompt into ChatGPT, Claude or whichever assistant you use, then replace every {{bracketed}} part with your own detail. The more specific and messier your input, the better the output — a model given raw notes has more to work with than one given a tidy summary you wrote first.
More Compensation prompts
Open the full tool for all 87 prompts, the glossary, and the JD decoder.