Most job descriptions read like legal disclaimers — a wall of responsibilities and requirements, heavy on corporate language, light on what the job actually feels like or why someone would want it. Writing a better one from scratch takes time most business operators don’t have.
AI tools like ChatGPT can generate a solid job description draft in minutes — but only if you give it the right inputs. This guide walks through how to prompt for better output, what to add that AI can’t know, and how to review the result for common mistakes that repel good candidates.
The Prompting Framework
The quality of an AI-generated job description depends almost entirely on what you put into the prompt. A good prompt includes:
- Role title and type — “Full-time Operations Coordinator, not a management role”
- Company context — “E-commerce brand selling outdoor gear, $3M revenue, 8 employees, Shopify-based, use a 3PL”
- Key responsibilities — the 5–7 most important things this person will actually do day-to-day
- Hard requirements — non-negotiable experience, skills, or certifications
- Nice-to-haves — useful but not required
- Tone and culture — “direct, low-ego, operationally rigorous team”
- Compensation range — if you’re including it (which you should for transparency)
Example prompt: “Write a job description for a full-time E-commerce Operations Coordinator at a $3M Shopify outdoor gear brand with 8 employees. The role manages day-to-day fulfillment coordination with our 3PL, inventory reordering, CS escalations, and weekly ops reporting. Must have: 2+ years of e-commerce operations experience, Shopify familiarity, comfort with Excel/Google Sheets. Nice to have: experience with Extensiv or Cin7, Zoho experience. The team is lean, direct, and moves fast. We work remotely with weekly video check-ins. Salary range $55K–$70K. Include a brief company overview paragraph and a clear application instructions section.”
What AI Does Well
- Structures the document logically (overview, responsibilities, requirements, nice-to-haves, compensation, how to apply)
- Uses active verbs for responsibilities (“Coordinate daily,” “Manage weekly,” rather than “Responsible for”)
- Generates inclusive language when prompted (or avoids unnecessarily gendered language by default)
- Produces a readable, professional tone
- Generates multiple variants if you ask (e.g., “write two versions — one more detailed, one shorter and punchier”)
What You Need to Add (That AI Can’t Know)
AI generates from patterns in language data — it can’t know what makes your company distinct. You need to add:
- Specific tools you actually use — “We use Extensiv for inventory, Zoho Desk for CS, and Slack for communication”
- What the first 90 days actually look like — “You’ll spend the first month shadowing our current ops setup and documenting processes before taking ownership”
- What good looks like in this role — what does success look like at 6 months?
- Honest culture signals — if you’re a fast-moving, high-accountability environment, say so — it attracts people who want that and filters out those who don’t
- Why this role exists now — “We’re hiring because we’ve outgrown founder-managed operations” is more compelling and honest than a generic opener
Common Mistakes to Avoid
Too Many Requirements
Research consistently shows — including Harvard Business Review’s analysis — that long requirement lists drive away qualified candidates, particularly women and underrepresented groups. Distinguish hard requirements (must-haves) from preferred qualifications clearly, and keep the hard list short. If you have 10 “required” items, you’ll get fewer and narrower applications.
Generic Responsibilities
“Manage operations” tells a candidate nothing. “Coordinate daily order fulfillment with our 3PL partner, process returns, and maintain weekly inventory reporting” is specific enough that the right person can picture doing it.
No Compensation Range
Omitting salary signals that you either don’t know what the role is worth or you’re hoping to underpay. Many qualified candidates skip roles without a range. Publishing a range is increasingly standard — and in several Canadian and US jurisdictions, required by law.
Editing AI Language Into Corporate Blandness
If you get a good AI draft and then edit out all the direct language in favor of corporate buzzwords, you end up with the same forgettable JD you started with. Push in the direction of more specific and human, not more formal.
Frequently Asked Questions
Can AI write job descriptions?
Yes — with a detailed prompt (role title, company context, responsibilities, requirements, compensation, tone), AI generates well-structured drafts quickly. The draft needs human editing to add company-specific details — actual tools, culture signals, what success looks like — that AI can’t know.
What should a job description include?
Company overview, role summary, 5–7 specific responsibilities (active verbs), hard requirements, preferred qualifications, compensation range, work arrangement, and application instructions. Optionally: what success looks like at 6 months and culture signals.
Should I include a salary range in a job description?
Yes — increases application quality, reduces wasted interviews, and is legally required in several Canadian provinces and US states (California, New York, Colorado, and others). Signals transparency and attracts strong candidates.
How many requirements should a job description have?
Limit hard requirements to 4–6. Long lists drive away qualified candidates. Clearly distinguish requirements from preferred qualifications so the right people self-select in rather than filtering out due to one or two nice-to-haves.
Build the Operations Team You Actually Need
OpsStack helps growing businesses define their operations roles clearly — job descriptions, org design, and onboarding SOPs that set new hires up for success. Talk to us about your next hire.
