Skip to playbook

HR06·Offer & Negotiation

Offer Email & Negotiation Talking Points

Draft the offer email and negotiation talking points from interview recommendations, compensation inputs, and hiring-risk constraints.

Before you start

What you’ll need

  • Completed interview debrief with a clear hire recommendation from HR05
  • Background check cleared or risk summary available from HR07
  • Approved compensation band for the target role
  • Compensation Guidelines
  • Candidate Expectations Notes

Who’s involved

  • Recruiter — Drafts the offer email and compiles negotiation talking points for the hiring manager.
  • Hiring Manager — Reviews offer terms, leads candidate negotiation, and confirms final acceptance details.

Safe use

  • Data Handling — This workflow processes internal compensation bands, candidate salary expectations, and background check findings. Do not paste exact salary figures or confidential background check details into unapproved GenAI tools.
  • Verification — GenAI may hallucinate benefits, misstate role scope, or invent compensation terms. Verify every claim in the draft offer email against the approved requirements and compensation guidelines before sending.

Execution steps

Hybrid

Inputs

  • Confirm the interview debrief recommends proceeding with an offer
  • Confirm background check findings are cleared or risks are documented
  • Confirm the approved compensation band is current and authorized
  • Confirm candidate expectations have been captured from recruiter conversations

Data Handling: Use only the approved compensation range in the prompt; do not include the full internal salary band or grading structure.

Prompt

Draft structured offer email from verified inputs

CONTEXT
You will be provided with the following source documents:
1. Interview Debrief Summary
2. Background Check Risk Summary
3. Compensation Guidelines
4. Candidate Expectations Notes

TASK
Draft a professional offer email addressed to the candidate. The email must state the role title, start date placeholder, compensation within the approved range, key benefits, and any conditions. Use a welcoming but direct tone.

OUTPUT FORMAT
Produce the email in markdown with the following sections:
- **Subject Line**
- **Greeting** (use [Candidate Name] placeholder)
- **Role & Team Overview** (2–3 sentences summarizing the role and team)
- **Compensation & Benefits** (bullet list of key terms within approved range)
- **Conditions & Next Steps** (any contingencies, response deadline placeholder)
- **Closing**

EXAMPLE
**Subject Line:** Your Offer for [Role Title] at [Organisation Name]

**Greeting:** Dear [Candidate Name],

We are pleased to extend an offer for the position of [Role Title] within the [Team/Department] team...

**Compensation & Benefits:**
- Base salary: [Amount within approved range]
- Annual leave: [X days]
- Benefits: [Summary from Compensation Guidelines]

**Conditions & Next Steps:**
- This offer is contingent on [any conditions from risk summary].
- Please respond by [Date Placeholder].

CONSTRAINTS
Do not include specific salary figures beyond the approved range provided. Do not invent benefits or perks not listed in the Compensation Guidelines. Do not include confidential background check details in the email body.

Outputs

  • Confirm the role title matches the approved position from the debrief
  • Confirm compensation terms fall within the approved range
  • Confirm no fabricated benefits appear in the email

Verification: Verify the AI did not fabricate benefits, misstate the role title, or add compensation terms not present in the guidelines.

GenAI

Prompt

Generate manager negotiation talking points

CONTEXT
You will be provided with an Interview Debrief Summary (including the hire recommendation and concerns), Compensation Guidelines (approved range and flexibility), and Candidate Expectations Notes (what the candidate has indicated they value).

TASK
Generate a set of negotiation talking points for the hiring manager. Each point should anticipate a likely candidate question or counter-offer scenario and provide a recommended response strategy within the approved parameters.

OUTPUT FORMAT
Produce a markdown document with:
- **Candidate Profile Summary** (2–3 sentences on what the candidate values most)
- **Talking Points Table:**

| # | Anticipated Concern | Recommended Response | Flexibility Boundary |
|---|---|---|---|
| 1 | [Concern] | [Response strategy] | [What can/cannot be adjusted] |

Include at least four talking points covering: base compensation, benefits, start date, and role scope.

- **Escalation Triggers** (bullet list of scenarios where the manager should consult HR before responding)

CONSTRAINTS
Do not suggest concessions outside the approved compensation range. Do not reference specific candidate personal details beyond what is in the Expectations Notes. Do not assume the candidate will negotiate — frame points as preparations, not predictions.

Outputs

  • Confirm all suggested concessions fall within approved flexibility limits
  • Confirm at least four distinct talking points are generated
  • Confirm escalation triggers are realistic and actionable

Verification: Verify the AI did not suggest concessions outside the approved compensation band or invent flexibility that does not exist.

Hybrid

Inputs

Prompt

Cross-check offer artifacts for accuracy

CONTEXT
You will be provided with a Draft Offer Email, Draft Negotiation Talking Points, and the original Compensation Guidelines.

TASK
Compare each claim in the offer email and talking points against the Compensation Guidelines. Flag any discrepancies: incorrect salary ranges, unlisted benefits, flexibility boundaries that exceed approved limits, or missing conditions.

OUTPUT FORMAT
Produce a markdown checklist:
- **Offer Email Review:**
  - [ ] Salary range matches approved band: [Yes/No with detail]
  - [ ] Benefits listed are accurate: [Yes/No with detail]
  - [ ] Conditions are complete: [Yes/No with detail]
  - [ ] Role title and scope are accurate: [Yes/No with detail]

- **Talking Points Review:**
  - [ ] Flexibility boundaries within approved limits: [Yes/No with detail]
  - [ ] Escalation triggers are appropriate: [Yes/No with detail]

- **Issues Found:** [List any discrepancies, or "None" if clean]

CONSTRAINTS
Do not suggest changes to the compensation guidelines themselves. Only flag mismatches between the drafted artifacts and the source guidelines. Do not add new offer terms.

Outputs

  • Verified Offer Pack
  • Confirm all salary and benefits claims match the approved Compensation Guidelines exactly
  • Confirm no background check details have leaked into the offer email
  • Confirm talking points do not promise concessions beyond approved flexibility

Verification: Verify the AI flagged all genuine mismatches and did not incorrectly approve claims that differ from the source guidelines.

Human

Inputs

  • Verified Offer Pack
  • Confirm the hiring manager has reviewed and approved the final offer email
  • Confirm negotiation talking points have been shared with the hiring manager
  • Confirm all flagged discrepancies from the verification step have been resolved

Outputs

  • Sent Offer Record
Hybrid

Prompt

Compile accepted offer details for onboarding handoff

CONTEXT
You will be provided with the Sent Offer Record (the final offer terms as accepted by the candidate) and Candidate Expectations Notes (preferences the candidate communicated during the hiring process).

TASK
Compile an Accepted Offer Details document that summarizes the agreed terms and candidate context needed by the onboarding team.

OUTPUT FORMAT
Produce a markdown document with:
- **Candidate Name:** [Placeholder]
- **Role Title:** [As offered]
- **Start Date:** [Agreed date]
- **Compensation Summary:** (base, benefits, any negotiated adjustments)
- **Special Conditions:** (any agreed accommodations or deferred items)
- **Candidate Context for Onboarding:** (2–3 sentences on what matters to this candidate)
- **Outstanding Items:** (anything requiring follow-up before day one)

CONSTRAINTS
Do not include internal negotiation strategy or hiring manager notes in the handoff document. Only include terms the candidate has explicitly agreed to. Do not add speculative context about the candidate beyond what is documented.

Outputs

  • Confirm all accepted terms match the final offer as sent
  • Confirm no internal negotiation notes appear in the handoff document
  • Confirm the document contains all fields needed by the onboarding workflow

Reference

Guardrails

  • Approved Range Only — All compensation figures in AI-generated artifacts must fall within the pre-approved band — never prompt the AI with or allow output beyond authorized limits.
  • Source-Backed Claims — Every benefit, condition, and role detail in the offer email must trace to a specific line in the Compensation Guidelines or Interview Debrief Summary.
  • Candidate Data Minimization — Include only the candidate details necessary for each prompt — exclude personal identifiers, current employer, or sensitive background findings from AI inputs.

Pitfalls

  • Pasting the full internal salary band structure into the AI prompt instead of the candidate-specific approved range.
  • Relying on AI-generated benefits descriptions without verifying them against your organisation's current benefits documentation.
  • Copying AI-drafted negotiation talking points verbatim without adapting them to the specific candidate's stated priorities.
  • Including confidential background check findings in the offer email prompt context, risking data leakage.

Definition of Done

  • The offer email accurately reflects the approved compensation range and all role requirements from the interview debrief.
  • Negotiation talking points cover at least four anticipated candidate concerns with response strategies within approved limits.
  • The Accepted Offer Details document contains all fields required by the onboarding workflow.
  • No confidential salary bands, internal grading, or raw background check data appear in any AI-generated output.

GET FULL ACCESS

Unlock every step, prompt, and downloadable example — for this playbook and the rest of AGASI AiOS, our GenAI capability framework.

We'll send a magic link — no password needed.

AGASI AiOS · HR06 v1.0 · Apr 7, 2026