Writing Job Descriptions That Attract the Right Talent

The Copy-Paste Problem

Here’s how most job descriptions get written: someone opens the last posting for a similar role, changes the title, tweaks a few bullet points, and hits publish.

The result? Descriptions that are generic, bloated with requirements nobody actually needs, and written in corporate jargon that makes candidates’ eyes glaze over. Worse, they often contain language that actively discourages diverse applicants from applying.

If your hiring pipeline isn’t producing diverse, qualified candidates, the problem might start at the very first touchpoint: the job description itself.

What You’ll Learn

By the end of this lesson, you’ll be able to write job descriptions that are inclusive by design, specific enough to attract the right people, and compelling enough to stand out in a market where top candidates have options.

From Lesson 1 to Lesson 2

Remember the sandwich model from Lesson 1? Job descriptions are a perfect application. You provide context about the role and your organization (top bread), AI generates a structured draft (filling), and you review for bias, accuracy, and cultural fit (bottom bread). Let’s put that into practice.

Why Most Job Descriptions Fail

Research from LinkedIn, Glassdoor, and Textio reveals consistent patterns in underperforming job postings:

Too many requirements. The famous Hewlett-Packard internal study found that women applied to jobs when they met 100% of qualifications, while men applied at 60%. Every unnecessary requirement is a filter that disproportionately excludes underrepresented candidates.

Biased language. Words like “rockstar,” “aggressive,” “dominant,” and “competitive” are associated with masculine-coded language and measurably reduce applications from women. Words like “supportive,” “collaborative,” and “nurturing” can reduce applications from men.

Vague responsibilities. “Drive business growth” means nothing. “Increase enterprise sales pipeline by managing 50+ accounts in the healthcare vertical” means something.

Missing information. Candidates want to know: What will I do every day? What do I need to succeed? What’s the compensation? What’s the team like? Missing answers mean missing candidates.

Quick Check

Pull up the last job description your organization posted. Count the number of requirements listed. How many are genuinely must-haves versus nice-to-haves that could be learned on the job?

The Inclusive Job Description Framework

Here’s a structure that works:

1. The Hook (Why This Role Matters)

Start with impact, not company history. Candidates want to know why this job exists and why it matters.

Weak:

“Founded in 2010, Acme Corp is a leading provider of enterprise solutions with offices in 12 countries…”

Strong:

“Our customer success team is the reason clients stay. As a Customer Success Manager, you’ll own relationships with our 50 largest accounts – the ones that drive 60% of our revenue. When you do this well, clients grow with us for years.”

2. Responsibilities (What You’ll Do)

Use specific, outcome-focused language. Start each bullet with an action verb and include enough context that candidates can picture the actual work.

Weak:

• Manage projects
• Interface with stakeholders
• Drive results

Strong:

• Lead 3-5 concurrent implementation projects from kickoff to launch, coordinating across engineering, design, and client teams
• Run weekly status calls with client stakeholders, proactively identifying blockers before they become problems
• Own the project timeline and budget, escalating risks when projects are trending more than 10% over scope

3. Requirements (What You Need)

This is where most bias creeps in. Be ruthless about separating genuine requirements from preferences.

Rules for requirements:

• Only list what’s truly required on day one. If someone could learn it in the first 90 days, it’s not a requirement.
• Focus on skills, not pedigree. “3+ years of experience managing B2B customer accounts” instead of “MBA from a top-tier university.”
• Use ranges, not minimums. “3-5 years” is more specific than “5+ years” and doesn’t accidentally exclude someone with 4 years of excellent experience.
• Eliminate nice-to-haves entirely. Or clearly separate them with “Bonus but not required” language. Better yet, leave them out and let candidates surprise you.

Weak:

• Bachelor’s degree required; MBA preferred
• 7+ years of experience
• Must be a self-starter and team player
• Expert in Salesforce, HubSpot, Tableau, Excel, Python, SQL

Strong:

• 3-5 years of experience in B2B account management or customer success
• Track record of retaining and growing client relationships
• Comfortable with CRM tools (we use Salesforce, but tool-specific experience isn’t required)
• Clear written and verbal communication skills

4. What You Get (Benefits and Culture)

Candidates are evaluating you, too. Be specific about what makes your company worth working for.

Weak:

“Competitive salary and benefits.”

Strong:

• Salary range: $85,000-$105,000 depending on experience
• Fully remote with quarterly in-person team gatherings
• 20 days PTO + 10 company holidays + your birthday off
• $2,000 annual learning budget
• Medical, dental, vision for you and dependents

Using AI to Draft Job Descriptions

Here’s a prompt template that produces strong first drafts:

Write a job description for [Job Title] at [Company Type/Size].

Role context:
- This person will [primary purpose of the role]
- They'll report to [manager title] and work with [key collaborators]
- The team currently has [size/structure]

Key responsibilities (3-5):
- [Specific responsibility with context]
- [Specific responsibility with context]
- [Specific responsibility with context]

Must-have qualifications (3-4 maximum):
- [Genuine requirement]
- [Genuine requirement]
- [Genuine requirement]

Compensation: [Range]
Work model: [Remote/hybrid/onsite]

Instructions:
- Use gender-neutral language throughout
- Focus on outcomes and skills, not pedigree or credentials
- Write in second person ("you'll" not "the candidate will")
- Keep the total posting under 600 words
- Start with why this role matters, not company history

The key is giving AI specific context rather than generic instructions. The more you tell it about the actual role, the better the output.

Checking for Bias

After AI generates your draft, run this bias checklist:

Language check:

• Are there any gendered terms? (“manpower,” “chairman,” “guys”)
• Any aggressive/competitive language that might discourage some candidates? (“crush targets,” “dominate the market”)
• Any age-coded language? (“digital native,” “young and energetic,” “seasoned veteran”)
• Any ability-coded language? (“must be able to stand for 8 hours” when the job is desk-based)

Requirements check:

• Is every requirement genuinely needed on day one?
• Could any requirement be replaced with a skill-based alternative?
• Does the education requirement match the actual work? (Do you really need a degree, or do you need someone who can do the job?)
• Are years-of-experience requirements reasonable for the level?

Tone check:

• Would this appeal to someone from a different background than your current team?
• Is the writing accessible, or filled with industry jargon?
• Does it sound like a real human wrote it, or corporate boilerplate?

Quick Check

Try running one of your existing job descriptions through the Job Description Bias Checker skill. Note what it flags – you might be surprised by language you didn’t realize was exclusionary.

Common Mistakes and Fixes

Exercise: Rewrite a Job Description

Take this generic job description and rewrite it using the framework above:

Marketing Manager We are looking for an experienced marketing manager to join our growing team. The ideal candidate will have 8+ years of experience, an MBA, and expertise in digital marketing, content strategy, SEO, SEM, social media, email marketing, and analytics. Must be a self-starter who thrives in a fast-paced environment. Competitive salary and great benefits.

What to fix: Too many requirements, vague language, credential-focused, missing salary, missing what the person actually does.

Key Takeaways

• Job descriptions are your first filter – biased descriptions produce biased pipelines
• Lead with impact, not company history – candidates want to know why the role matters
• Limit requirements to genuine must-haves – every unnecessary requirement filters out diverse candidates
• Use AI to generate first drafts, then review specifically for biased language, inflated requirements, and vague responsibilities
• Include salary ranges and specific benefits – transparency attracts better candidates

Next lesson: you’ve attracted the candidates. Now let’s build a fair, efficient screening process.