Writing Sourcing Prompts for Recruiters: Get More Out of AI
Learn to write effective sourcing prompts with 10 practical examples for recruiters. Improve your AI prompts, LinkedIn Boolean searches and outreach.
Using AI effectively for sourcing requires a structured prompt with six fixed components to generate directly usable results. Specifying context, target audience and channel avoids generic output and saves a significant amount of time.
Writing sourcing prompts means directing AI with a short, clear briefing, so you get output for recruitment you can actually use straight away. You name the role, the context, the target audience, the channel and the output you want. This stops you getting vague answers, and gets you to concrete search strings and messages faster. In this article, you'll discover how to apply this in practice, including handy, copy-and-paste examples.
- A good prompt works like a mini-briefing with fixed components.
- Clear context leads to better AI output for sourcing and outreach.
- Short, focused prompts save time and lift response rates.
- Reusing and refining prompts keeps quality consistent.
Why writing sourcing prompts makes the difference
A lot of AI output stays very generic, because the question is often too broad. A simple instruction without context quickly leads to boilerplate text that delivers little. When writing a sourcing prompt, you give AI exactly the same information you'd hand to a colleague. Think of the role, sector, region and target audience. On top of that, add the channel and the tone you want. That's how you get output you can actually use for recruitment straight away.
In practice, we see that fixed AI sourcing prompts mean you have to rewrite far less. That doesn't just save time; it also helps you stay focused. The result is highly consistent communication with candidates.
What writing good sourcing prompts means in practice
A strong prompt has six fixed components that together give clear direction to the output. This structure works perfectly for general recruitment prompts, but just as well for prompts written specifically for a local audience.
- Role: who the AI should be, for example a senior recruiter.
- Context: the specific vacancy or assignment.
- Target audience: exactly who you're looking for, including the level you want.
- Channel: the platform you're using, such as LinkedIn, email or a job board.
- Tone: the writing style you want, for example direct, personal or informal.
- Output: the expected format, such as a maximum number of sentences or a fixed structure.
A short comparison makes the difference clear straight away: a vague instruction stays generic, while a specific instruction delivers a result immediately.
Weak prompt: create a Boolean search for a recruiter.
Good prompt: create a Boolean search prompt for LinkedIn for a nurse in the Leeds area at degree level, with hospital experience, and exclude internships. Give the string and three variations.
Working with fixed formats? You can reuse these easily through fixed prompt templates. This lets you build your own library of handy, ready-to-use recruiter prompts remarkably quickly.
Common mistakes when writing sourcing prompts
The most common mistake is giving too little context, which leaves the AI not understanding exactly what the goal is. A clear brief on the output you want is also often missing. You then end up with, for example, endlessly long text, when what you actually needed was a short, punchy message. Another pitfall is using English unnecessarily when your target audience simply speaks another language. Local job titles and terms simply land far better.
On top of that, many recruiters don't refine their prompt after a first attempt. Good results often only appear after two or three iterations. Finally, pay close attention to privacy: never share personal data in public AI tools, and always carefully check facts, such as salaries and regulations, yourself.
Tip: Elvatix gets more out of every InMail credit. Higher response rates, lower cost per contact.
See how →How to apply writing sourcing prompts in your daily workflow
Always start with the intake, and sum up the essence of the role in a few sentences. Then turn this into a clear sourcing briefing prompt, in which you name the role, target audience and channel. Use the output as your starting point for the search or outreach. Then critically check whether the tone and content genuinely fit your target audience. Not quite right? Adjust the prompt and simply try again. Both AI-generated screening questions and outreach prompts for recruiters get considerably better the more you keep refining them.
Make working with prompts a fixed part of your process, and keep successful variants stored in an organised way. On the page about how this fits into your daily work, you can see exactly how to build this approach into one streamlined workflow.
10 examples for writing sourcing prompts you can use right away
Use the examples below as a strong starting point. Adjust them where needed per sector, so they fit your specific target audience seamlessly.
Prompt 1: Boolean string for LinkedIn
Use case: actively searching through LinkedIn Recruiter.
Good prompt: Create a Boolean search prompt for LinkedIn for a maintenance technician in the Manchester area with a Level 3 or 4 vocational qualification and experience in production, excluding freelancers. Give one main string and three variations for the engineering, logistics and food sectors.
Example output: ("maintenance technician" OR "maintenance engineer") AND (NVQ OR "level 3" OR "level 4") AND (production OR factory) NOT (freelance OR contractor)
Prompt 2: Outreach message for passive candidates
Use case: first contact via LinkedIn.
Good prompt: In three sentences, write an outreach message for LinkedIn aimed at a data analyst in London. Name a concrete project, keep the tone personal, and close with an open question.
A clear structure considerably lifts response rates. In this real-world example of a strong InMail approach, you can see nicely how structure and personalisation come together successfully.
Prompt 3: Generating screening questions
Use case: preparing an intake.
Good prompt: Create five screening questions based on this specific vacancy. Focus on hard requirements, recent work experience and availability. Phrase short questions that fit well into a first phone introduction.
Prompt 4: Inclusive job ad copy
Use case: appealing to a broader audience.
Good prompt: Rewrite this job ad with a prompt focused on inclusivity. Use plain, accessible language, avoid gendered wording, and keep the sentences short and punchy.
Prompt 5: Estimating the salary range
Use case: preparing for a conversation.
Good prompt: Give a reliable salary estimate for a nurse with five years' work experience in the Birmingham area. Explain the chosen range clearly and name any assumptions made. Use the most recent market data where possible.
Prompt 6: Building a candidate profile
Use case: sourcing in a more targeted way.
Good prompt: Create a candidate profile prompt for a senior engineer. In five sentences, describe the motivation, working environment, possible reasons to move, and ideal projects.
Prompt 7: Intake to briefing
Use case: bringing structure to the information.
Good prompt: Smoothly turn this intake into a sourcing briefing prompt, including must-haves, nice-to-haves, target audience and search strategy. Present the result in five short, clear bullet points.
Prompt 8: Rejection email
Use case: improving the candidate experience.
Good prompt: Write a personal rejection email with concise feedback, and directly ask for consent to be added to the talent pool. Keep the tone respectful and clear throughout.
Prompt 9: Talent mapping
Use case: gaining in-depth market insight.
Good prompt: Create a talent-mapping prompt for software testers in the Bristol area. Name relevant companies and job titles, and briefly explain why these companies are such a good fit.
Prompt 10: Analysing outreach
Use case: raising conversion.
Good prompt: Analyse 100 LinkedIn messages and give targeted improvement points for the opening lines, personalisation and call-to-action. Name three concrete changes to make.
How to build a strong prompt library by writing sourcing prompts
Save your very best prompts in a structured way, by role and by channel. Give each prompt a clear name, and carefully track what does and doesn't work. Measure the response, and adjust your prompts based on the results you get. That's how a practical library of highly effective recruitment prompts builds up quickly. Working as a team? Then agree on fixed formats and a consistent tone of voice. That leads to recognisable, professional communication. Read more about working with prompts as a team and how to apply these agreements consistently.
Download it and try it yourself with writing sourcing prompts
Bundle all your prompts neatly into one organised document, for example in Notion or Markdown. Use clear categories, such as AI sourcing prompts, screening questions and outreach. On top of that, add variants per sector, so you can always switch quickly. Want to test this directly in your own workflow? Then head to try it yourself and see straight away how better prompts lead to considerably higher-quality output.
Working safely and smartly with AI in recruitment
Always use AI as a tool, and critically check the content. Don't share sensitive personal data, and always double-check important information, such as current salaries and regulations. Writing good sourcing prompts ultimately comes down to the right balance. You give AI enough context for direction, while keeping full control over the quality of the final result.
Frequently asked questions about writing sourcing prompts
What is writing sourcing prompts, in one sentence?
It's formulating a short, powerful briefing for AI, including the role, context, target audience, channel and desired output, so you get recruitment results you can use straight away.
How long should a good prompt be?
Keep it compact above all. One to three sentences is usually more than enough, as long as all the important components are present in the briefing.
Should I write prompts in my candidates' own language?
Yes, definitely when your target audience speaks that language. Prompts written in the local language fit local job titles far better, and raise the relevance of your search.
How do I improve my prompts?
Test different variants, compare the output, and adjust deliberately on tone, context and the format you want. Continuous iteration is guaranteed to keep delivering better results.
Is AI reliable for salary and market data?
Use AI as a first indication at most, and always check the data against reliable, current sources. So never make final decisions without validating this data yourself.
Write a personal message right here
Enter a name or LinkedIn URL and get a personalised message within 30 seconds. No account needed.
Optimise your recruitment workflow with AI and Elvatix?
Elvatix helps you structure your sourcing process with smart templates and AI integrations. Manage your prompts centrally for consistent, high-quality candidate communication.
