Overview
This article helps you use the Bias Analyser in Access Volcanic to create more inclusive job adverts. The tool examines your advert content for gender, cultural, age, and experience bias, providing scores and actionable suggestions to help you attract a broader, more diverse candidate pool.
Developed in partnership with the University of Nottingham's Applied Linguistics Department, the Bias Analyser uses linguistic research and AI to identify problematic language and structural issues that might discourage qualified candidates from applying.
Key benefits
Using the Bias Analyser can help you:
Attract more diverse candidates – Research shows optimised adverts receive 68% more female applicants and 28% more applicants overall
Improve candidate quality – Inclusive adverts double the number of relevant applicants shortlisted
Enhance your employer brand – Demonstrate your commitment to diversity, equity, and inclusion
Boost search performance – LinkedIn's algorithm recommends optimised profiles 30% more frequently to female candidates
📌 Note: These performance improvements are based on internal research and external studies conducted in partnership with the University of Nottingham.
Before you start
Before using the Bias Analyser, make sure you have:
Access to the Access Volcanic admin area
Permission to view the Jobs section and Dashboard
A job advert ready to analyse (either new or existing)
📌 Note: The Bias Analyser is enabled by default for all Access Volcanic customers.
Understanding what the tool analyses
The Bias Analyser examines two key areas of your job adverts:
Language score
This score measures how inclusive and welcoming your wording is. The tool identifies:
Gender bias – Masculine-coded terms like champion, driven, or strong
Cultural bias – Phrases like English native speakers or references to top-tier universities
Age bias – Terms that suggest age preferences or physical demands
Experience bias – Language that unnecessarily excludes less experienced candidates
Tone issues – Overly formal or impersonal language that creates barriers
Structure score
This score evaluates how well your advert is organised and whether it includes essential information:
Length and layout – Ideal adverts contain 400 to 750 words with clear paragraphs, subheadings, and bullet points
Salary and benefits – Transparency about compensation
Location and working pattern – Clear details about where and when people work
Company culture – Information that helps candidates understand your team
Key responsibilities – The top three to five duties of the role
Required skills – The top three to five must-have qualifications or experience
Application process – Clear steps and timelines
Equal Opportunity statement – An authentic, conversational statement explaining your commitment to diversity
Running the Bias Analyser
Analysing your job advert
Log in to Access Volcanic as an admin
Navigate to the Jobs section
Create a new job advert or select an existing one to edit
In the Job Description field, type or paste your advert text
Click the Run Bias Analyser button at the top-right of the content box
The editor will switch to analysis mode – click Review Ad to begin
Wait a few moments whilst the tool analyses your content
Review your two scores (both out of 100):
Language score – measures inclusive language and tone
Structure score – assesses content completeness and readability
Understanding your results
What the scores mean:
70 to 80 or higher – Strong score, your advert is inclusive and well-structured
Below 70 – Improvements needed to attract a diverse candidate pool
The tool highlights biased phrases using colour codes and underlines.
🤓Tip: Hover over the purple tooltip icons to understand why specific terms are flagged and see suggested alternatives.
Replacing biased terms
Click on any highlighted word in your advert
Read the explanation of why the term is considered biased
Review the inclusive alternatives suggested by the tool
Click your preferred replacement to update the text instantly
Click Review Ad again to refresh your scores and see improvements
Common biased terms and alternatives
Biased term | Inclusive alternative |
Champion | Advocate, Promote, Support |
Driven | Motivated |
Strong | Experience with |
Work hard, play hard | Work smart, play more |
Top-tier universities | Relevant experience |
English native speakers | Fluent English speakers |
He/She | You |
Improving your structure score
If your structure score is low, consider adding or strengthening these elements:
Include essential information
Make sure your advert clearly states:
Salary range or benefits package
Office location or remote working options
Working hours and flexibility
Team size and reporting structure
Application deadline and interview process
Write an effective Equal Opportunity statement
Your Equal Opportunity (EO) statement should:
🤓Tip: Place your EO statement in the top half of your advert where candidates will see it early.
Do | Avoid |
Use clear, conversational language | Generic phrases like: We are an equal opportunity employer |
Explain why diversity matters to your organisation | Heavy legal wording that feels like a disclaimer |
Cover key employment areas (hiring, training, pay, promotion) | Simply listing protected characteristics without context |
Show commitment to continuous improvement | Leaving it at the bottom of your advert as an afterthought |
Align with your company values authentically | Letting it become outdated – review and update regularly |
Include practical support details where relevant |
|
📌Learn more: For more examples and guidance, read AIHR's 6 Inspiring Equal Opportunity Employer Statement Examples [+ Guide to Craft Your Own] guide.
Working with formatting
Best practice for formatting
The Bias Analyser works most accurately with plain text. For best results:
Write your advert using simple formatting – headings, bullets, and paragraphs
Avoid colours, tables, images, or complex HTML whilst analysing
Run the Bias Analyser and make your improvements
Click Back to Formatting to reapply your visual styles afterwards
📌Note: If you run the analyser on heavily formatted content, the highlighting may clash with your existing colours, or structured elements like tables might be misinterpreted. The tool will not warn you about this, so start with plain text for accurate results.
Using the tool with external systems
Integrated or imported jobs
The Bias Analyser does not run automatically on jobs imported from CRM systems or multi-poster integrations.
To optimise imported jobs:
Open the job within Access Volcanic
Run the Bias Analyser and make improvements
Copy the optimised text
Paste it back into your CRM or external system
This ensures your improved content appears everywhere you advertise the role.
Beyond job adverts
Whilst the Bias Analyser is designed for job adverts, you can also use it to improve:
Website copies
Marketing emails
Candidate communications
Internal announcements
📌Note: When analysing non-job content, only the Language score will be meaningful. The Structure score specifically evaluates job advert requirements.
Best practices
Creating inclusive job adverts
Start with plain text for accurate analysis
Aim for 400 to 750 words with clear paragraphs and bullet points
Include all essential information upfront (salary, working pattern, responsibilities, application process)
Write your Equal Opportunity statement authentically, explaining why inclusion matters to your organisation
Review and update your EO statements regularly to keep them current
Replace masculine or biased terms consistently across all adverts
Use you instead of gendered pronouns throughout
Optimising your scores
Focus on the suggestions with the biggest impact first
Do not chase a perfect score of 100 – anything above 70 to 80 is strong
Make content improvements (not just word swaps) to boost your structure score
Re-run the analyser after significant changes to track your progress
FAQs
Q1: Do I need to achieve a score of 100?
Answer: No, a score of 70 to 80 or higher is considered strong and inclusive. Aiming for perfection is not necessary to create an effective, bias-free job advert.
Q2: Why did my structure score not improve after making changes?
Answer: Structure improvements require adding missing content, not just replacing words. Make sure your advert includes salary information, a strong Equal Opportunity statement, clear responsibilities, and essential details about the role and application process.
Q3: Why has my formatting changed after using the tool?
Answer: The analyser temporarily adds colour highlights and underlines to flag biased terms. To avoid this, start with plain text, complete your optimisation, then click Back to Formatting to reapply your styles.
Q4: Can I turn off the Bias Analyser?
Answer: The Bias Analyser is enabled by default for all Access Volcanic customers and available to all users. Using the tool is entirely optional – it will not analyse your adverts unless you actively navigate to it from your job description and select Review Ad. There is no way to disable the feature through the platform. If you feel strongly about this, contact your Account Manager or Customer Success Manager to discuss your concerns.
Q5: Can I use the Bias Analyser for content that is not a job advert?
Answer: Yes, but only the Language score will be relevant for non-job content. The Structure score is designed specifically for job adverts and evaluates elements like salary information and Equal Opportunity statements.
Q6: Which languages do the tool support?
Answer: The Bias Analyser supports English only. The research and datasets used to build the tool are based on English-language job adverts, so analysing content in other languages will not produce meaningful results.
Q7: Does the tool analyse jobs automatically?
Answer: No, you need to run the Bias Analyser manually for each job advert. It does not run automatically on new or imported jobs.
Q8: Who can access the Bias Analyser?
Answer: Anyone with access to the admin area, specifically the Jobs section and Dashboard, can use the Bias Analyser. This does not require full admin rights – custom admin groups with these permissions can also use the tool.