ChatGPT Prompts for Recruiters: 18 Templates Across the Hiring Funnel

Most “200 ChatGPT prompts for recruiters” lists are unusable. Vague, generic, and missing the one thing that matters: context. A prompt that says “write a job description for a marketing role” produces marketing-role boilerplate. The same prompt with role specifics, audience, and output structure produces something a hiring manager will actually approve.

Johannes Sundlo’s prompt walkthrough takes the opposite approach. Instead of dumping a 200-prompt PDF, he walks one hire (a Social Media Manager role) through ChatGPT one stage at a time: JD, sourcing string, pre-screening questions, interview templates, outreach messages. The framework is what matters. Stage the prompts. Keep each one specific. Let each output feed the next.

The 18 prompts below are ones we use at Recrudoc, built on Sundlo’s staged approach. They’re not from his videos. He shows the framework with a handful of brief examples; we’ve written out fillable templates for every stage of the funnel. Use them as starting points and adapt to your roles.

How to use this library

In short: Save these as snippets in a notes app or text expander. Adapt the bracketed sections per role. Run them in sequence so each prompt’s output feeds the next. The more specific your input, the better the output, every time.

Two rules that apply to every prompt below:

1. Specificity beats cleverness. Johannes is direct on this: “The more specific I am, the better results from ChatGPT.” A two-line prompt with role title, seniority, company stage, location, and tech stack outperforms a 10-line prompt full of “act as a senior recruiter” filler.

2. Run prompts in sequence, not as one mega-prompt. “It’s better to do it separately versus doing one long,” Johannes notes. The JD prompt produces a JD. That JD becomes input to the search-string prompt. The search results inform the pre-screening prompt. Sequential prompts produce sharper outputs than one mega-prompt asking for everything at once.

Now the prompts, organized by funnel stage.

Stage 1: writing the job description

In short: Three prompts cover the JD stage: full JD generation, JD rewrite for specificity, and JD-to-LinkedIn-post conversion. Always feed ChatGPT the company context, role specifics, and tone constraints.

Prompt 1: full JD generation

Write a job description for a [TITLE] role at [COMPANY], a
[STAGE / SIZE / INDUSTRY] company.

Context:
- Reports to: [MANAGER]
- Location and remote policy: [LOCATION]
- Salary range: [RANGE]
- Top 3 things this hire needs to deliver in year one: [GOALS]

Output structure:
1. 2-sentence company hook (specific, not generic)
2. Role overview (3 sentences max)
3. Responsibilities (5-7 bullets, action verbs)
4. Requirements split into "Required" and "Preferred"
5. Benefits (real, not aspirational)
6. Application process

Tone: confident, specific, no clichés like "rockstar" or
"fast-paced". Length: 400-600 words. Use "you" not "the candidate".

When to use: starting a JD from intake notes or a template.

Prompt 2: JD rewrite for specificity

This JD is too generic. Rewrite it preserving the role and seniority
but replace every cliché with concrete details. Add team size, tech
stack, or business context where relevant. Cut filler.

Original JD: [PASTE]

When to use: a hiring manager hands you a generic copy-paste from an old role.

Prompt 3: JD to LinkedIn post

Convert this JD into a LinkedIn post that drives applications.

Format: hook line (no emoji), 3-line role summary, 4-5 bullets on
what the candidate gets, clear application CTA, 5 hashtags.
Length: under 1300 characters. Tone: human, not corporate.

JD: [PASTE]

When to use: every time you publish, since LinkedIn copy needs different structure than the JD itself.

For a full deep-dive on JD-writing prompt workflows, see our guide to AI job description prompts.

Stage 2: sourcing strings and search strategy

In short: ChatGPT can generate Boolean strings quickly, but the bigger value is creative sourcing brainstorms. The model is willing to suggest places to look beyond LinkedIn that you might not think of on your own.

Prompt 4: generate Boolean search string

Below is a job description. Generate:
1. A list of all technical and functional skills mentioned (one per line)
2. A narrow Boolean search string for LinkedIn Recruiter (high precision)
3. A broad Boolean search string for LinkedIn Recruiter (more candidates)

Use AND, OR, parentheses. Quote multi-word phrases.

JD: [PASTE]

When to use: every new role. Pair it with our LinkedIn Boolean search guide to sanity-check the output.

Prompt 5: Boolean variants by seniority

Take this Boolean string and produce 3 variants:
1. Junior version (1-3 years, removes senior-only signals)
2. Mid version (3-5 years)
3. Senior version (5+ years, adds leadership signals)

Original Boolean: [PASTE]

When to use: when the same role has multiple openings at different levels.

Prompt 6: creative sourcing brainstorm

I'm hiring a [ROLE] in [LOCATION/REMOTE]. Beyond LinkedIn and the
usual job boards, brainstorm 25 creative places to find candidates
for this role. Include niche communities, conferences, podcasts,
Slack/Discord groups, GitHub-style platforms, alumni networks,
and adjacent industries to recruit from.

Format: source name, why it works for this role, how to access it.

When to use: when LinkedIn is over-fished and you need fresh sources. Johannes points out that there’s no real limit on how many suggestions ChatGPT will produce. Asking for 25 instead of 5 produces a wider net at the same cost.

For sourcing strategies that don’t lean on prompts, see candidate sourcing strategies for 2026.

Stage 3: screening and filtering applicants

In short: Prompts in this stage shape your pre-screening questionnaire, the screening call script, and the rubric you’ll evaluate candidates against. Good screening prompts pay back across every candidate in the funnel.

Prompt 7: pre-screening questionnaire

Generate 10 pre-screening questions for a [ROLE] candidate.

Mix:
- 3 hard-skill questions (yes/no or short answer, fact-checkable)
- 3 experience questions (years, scope, tools used)
- 2 motivation/fit questions (open-ended, 1-2 sentences)
- 1 logistics question (location, salary expectation, notice period)
- 1 deal-breaker question (e.g. willingness to be on-call)

Format: numbered questions, each with the type label in parentheses.

When to use: building application forms or initial outreach questionnaires.

Prompt 8: recruiter screening call script

Build a 25-minute recruiter screening call script for a [ROLE]
candidate. Include:
- 2-minute intro (company pitch, role context)
- 12 minutes of candidate questions covering experience,
  motivations, and 2-3 must-have requirements
- 5-minute candidate Q&A
- 6-minute close (next steps, timeline, compensation expectations)

Format: timestamped sections with exact wording where useful.

When to use: standardizing screening calls across a team. For a deeper template library, see our screening call scripts guide.

Prompt 9: resume evaluation rubric

Create a resume evaluation rubric for a [ROLE] role.

Categories:
- Required skills match (0-5)
- Years of relevant experience (0-5)
- Career trajectory (0-5)
- Domain/industry relevance (0-5)
- Red flags (negative score, list each)

For each category, define what scores mean. Total out of 20.
Anything under 12 is auto-reject. 12-15 needs screening call.
16+ is fast-track.

Reference JD: [PASTE]

When to use: when multiple recruiters or coordinators are reviewing applicants and you need consistent decisions.

Stage 4: interview stage templates

In short: ChatGPT shines at generating role-specific interview question banks. The real leverage is asking for stage-specific templates separately (recruiter screen, hiring manager interview, technical interview, cultural fit, exec final) instead of one mega-list.

Johannes is explicit on this: he uses different prompts for the different stages and finds it yields the best results, better than running one long prompt.

Prompt 10: hiring manager interview template

Generate a 45-minute hiring manager interview template for a [ROLE].

Structure:
- 5-min intro and rapport
- 25-min experience deep-dive (3 deep questions on past projects,
  STAR-format probes for each)
- 10-min role-specific scenario question
- 5-min Q&A and close

Output: timestamped sections with the exact questions and
recommended follow-up probes for each.

Prompt 11: technical or skills interview

Build a 60-minute technical interview for a [ROLE].

Include:
- 1 take-home component (15-min ramp-up explanation if used)
- 1 live problem-solving exercise (30 minutes, with hints if stuck)
- 1 system/process design discussion (15 minutes)
- 5-minute candidate Q&A

For each section: the exact prompt to give the candidate, what
"strong" looks like, what "weak" looks like.

Prompt 12: cultural / values interview

Generate 8 values-based interview questions for a [ROLE] at a
company whose stated values are: [VALUE 1, VALUE 2, VALUE 3].

Each question should:
- Probe a specific value
- Use STAR (situation, task, action, result) format
- Have a clear "strong answer" and "weak answer" rubric

Format: question + value being tested + rubric.

Prompt 13: interview training guide for hiring team

Create an interview training guide for a [ROLE] hiring team that
has not interviewed for this kind of role before.

Cover:
- The 3 most important things to assess
- Common biases that show up in this role's interviews
- 2 behavioral red flags specific to this role
- 5 questions every interviewer in the loop should ask
- How to write good interview feedback (with a 1-paragraph example)

Length: 600-800 words. Tone: practical, not academic.

When to use: onboarding new interviewers. Sundlo flags this in his walkthrough as a useful application of ChatGPT for recruiting teams that need to prepare hiring panels for a role they haven’t run before.

Stage 5: candidate outreach and communication

In short: Outreach prompts tend to produce good drafts you’ll edit lightly. The leverage is volume. Drafting personalized messages with ChatGPT is meaningfully faster than writing each one from scratch.

Prompt 14: cold outreach message

Write a LinkedIn cold outreach message to a [TITLE] currently at
[COMPANY], for a [TARGET ROLE] at [HIRING COMPANY].

Constraints:
- Under 600 characters
- One specific, personalized opener referencing something concrete
  about their profile (tell me what to look for if you don't know it)
- One sentence on the role hook
- Clear, low-friction CTA (e.g. "open to a 20-min call?")
- No "I came across your profile" or other clichés

Output: the message + a 1-line note on what personalization element
the recruiter needs to fill in before sending.

Prompt 15: follow-up sequence (3 messages)

Write a 3-message follow-up sequence for a [TARGET ROLE] candidate
who didn't reply to the initial outreach.

Cadence:
- Message 2: 5 days after message 1, different angle, shorter
- Message 3: 10 days after message 2, break-up message, low pressure

Each message under 500 characters. No guilt, no "circling back",
no "just bumping this".

For a full breakdown of outreach message types, see our guide to recruiting messages in one click.

Prompt 16: rejection message (specific reason)

Write a candidate rejection message for someone who interviewed
but wasn't selected. Reason: [SPECIFIC REASON, e.g. "not enough
direct payments experience"].

Tone: respectful, specific without being harsh, leaves the door
open for future roles. Under 150 words. No "we've decided to move
forward with another candidate" boilerplate.

When to use: post-interview rejections where the candidate deserves better than a template.

Stage 6: process improvement and reporting

In short: Two prompts handle the meta-work of recruiting: analyzing your funnel and generating progress reports for hiring managers. Both turn data dumps into structured insight quickly.

Prompt 17: funnel analysis

I have hiring funnel data for the [ROLE]:
- Sourced: [N]
- Replied: [N]
- Phone screens: [N]
- Hiring manager interviews: [N]
- On-site/final: [N]
- Offers: [N]
- Hired: [N]

Calculate stage-to-stage conversion rates. Compare against typical
benchmarks for this kind of role. Identify the 1-2 biggest leaks
in the funnel and suggest 3 concrete actions to address each.

When to use: weekly or end-of-search retrospectives.

Prompt 18: hiring manager status update

Write a 1-screen status update for the hiring manager on the
[ROLE] search.

Include:
- Headline (one sentence: are we on track or not, and why)
- Funnel snapshot: [INSERT NUMBERS]
- 2-3 strongest candidates: name, current stage, 1-line summary
- 2-3 risks or blockers
- Asks from the hiring manager (1-2 specific things you need
  from them this week)

Tone: confident, factual, no padding. Under 250 words.

When to use: weekly hiring manager updates.

Where the prompt library hits a ceiling

In short: Managing this many prompts in a notes app is fine for a single recruiter on light volume. At higher volume or with a team of recruiters, prompt management itself becomes the bottleneck. You context-switch between ChatGPT, your ATS, and your candidate database. Recruiting CRMs that bake the prompts into the workflow remove the management overhead.

The honest limitation of any “ChatGPT prompts for recruiters” library is the workflow surrounding the prompts:

• You copy intake notes into a JD prompt
• You copy the JD into a sourcing-string prompt
• You copy a candidate’s profile into a screening-question prompt
• You copy the screening notes into a follow-up prompt
• You copy the response into your ATS

Each copy-paste cycle takes time. Across many candidates and stages, that adds up to real friction every week. The prompts themselves are free. The workflow tax is real.

Tools like Recrudoc flip this by baking the prompts into the recruiting workflow itself. The JD Parser turns intake into structured requirements. The AI Message Writer covers all 9 outreach types in 3 tones, with no prompt copying. Instant Scorecards generate the same kind of structured output you’d get from a screening rubric prompt, tied directly to the candidate record. The cost works out to roughly a small usage-based fee per AI operation. For context, our AI cost comparison shows how this stacks up against a $20/month ChatGPT subscription you’re still copy-pasting from.

A rough decision tree from our own experience helping recruiters set this up:

Volume	Our recommendation
Light, working solo	ChatGPT plus this prompt library
Moderate, working solo	ChatGPT plus a structured snippet manager
High volume or a team of recruiters	A CRM with AI built in

Volume changes which approach saves you time. Below the threshold, prompts in a notes app are adequate. Above it, prompt management itself becomes the work.

For a broader look at the AI tools recruiters are actually using, see the best AI tools for recruiters in 2026.

Tired of managing a library of ChatGPT prompts across tabs? Try Recrudoc CRM free. JD parsing, scorecards, and 9 message types are built into the workflow with no prompt copy-paste required.

Sources

The insights in this article are based on the following industry expert discussion:

• “Boost Recruitment Success with 200 Free ChatGPT Prompts” — Johannes Sundlo, YouTube