Solutions
Discover "Me" · Work Personality
THE AI COACH THAT ACTUALLY GETS YOU.
Voice or text coaching built on psychology. For you, your team, or the candidates you place.
Hey Compono!
A coach that actually gets you.
Get 10 minutes free, then $15 a month. Cancel anytime.
Get Started ≫Roast My Job Ad
Paste your job ad, get it roasted. Buzzwords, salary dodges, vague duties and coded language, scored out of 100 and quoted straight back at you. Your ad never leaves your browser.
Your job ad
The charge sheet
Every charge quotes your own ad back at you. The roast is the diagnosis; the fix underneath is the part worth forwarding to whoever wrote it.
Cleared without charge
The failures are consistent: no salary information, a wall of clichés ("fast-paced environment", "wear many hats"), duties so vague they could describe any job, a requirements list only a unicorn could meet, and far more words about the company than about what the candidate gets. Each one measurably cuts the number and quality of applicants. This tool checks for all of them at once and shows you where your ad stands.
A bad job ad is rarely a writing problem. It is what an unstructured hiring process looks like from the outside: nobody agreed what the role actually needs, so the ad asks for everything and says nothing. Compono Hire starts earlier, defining what great looks like for the role, then matching candidates on work personality and culture fit against that definition, so the ad can say something specific and the shortlist can prove it. Lyre's used Compono Hire to scale from 4 to 70 people across five continents in two years without losing fit.
See how it worksHow it's calculated
The roast runs eight checks in your browser: recruiting clichés (the rockstars, ninjas and fast-paced environments), salary transparency (a real range beats a figure, a figure beats "competitive"), vague duties ("assist with", "other duties as required"), unicorn hunting (requirement overload and years-of-experience demands that contradict the seniority), gender-coded language, age-coded phrasing, readability, and how much of the ad is about the company instead of the candidate. The coded-language check uses the masculine and feminine word lists from Gaucher, Friesen and Kay (2011), the peer-reviewed research behind most gender-decoder tools; it reports a lean, not a verdict. Each check deducts from a starting score of 100, weighted by how much the evidence says it hurts response rates, and the total maps to a roast grade from Rare to Burnt to a Crisp. Everything runs client-side. Your ad is never uploaded, stored or seen by us.
New to the term? Read the plain-English definition of pay transparency laws in the HR Glossary.
Only here for the gender check? The standalone Gender Decoder runs the same published word lists and shows every matched term, both directions.
Common questions
Is it safe to paste a real job ad in here?
Yes. The analysis runs entirely in your browser using word lists and pattern matching. Nothing you paste is uploaded, stored or sent to Compono or anyone else. Close the tab and it is gone.
Should I put the salary in a job ad?
Yes. SEEK's own research puts numbers on it: job ads with a visible salary receive about 36% more applications, and 87% of candidates say knowing the salary before applying is extremely important. "Competitive salary" is read by candidates as "below market or we would have said". A genuine range also filters out people you were never going to agree terms with, which saves everyone a round of interviews.
What is gender-coded language in a job ad?
Words that research links to gendered self-selection. Gaucher, Friesen and Kay (2011) showed ads heavy in agentic words like "competitive", "dominant" and "determined" attract fewer women applicants, while communal words like "supportive" and "collaborative" lean the other way. Neither list is banned language; the problem is a strong lean you did not intend. The roast reports the lean and lets you decide.
How is the roast score worked out?
Every ad starts at 100. Each of the eight checks deducts points based on severity, with the heaviest weights on missing salary information and cliché density because they do the most damage to application rates. 85 or above grades Rare, 65 to 84 Medium, 40 to 64 Well-Done, and below 40 is Burnt to a Crisp. The score is a guide for improving the ad, not a scientific instrument, and the fixes matter more than the number.
.webp)
.png?width=383&height=200&name=team%20(1).png)