How to use an AI salary estimator to negotiate your next offer

How AI salary estimators actually work

Most estimators combine three data sources: crowdsourced self-reports from workers, structured salary data from job postings that include pay bands, and government filings such as H-1B Labor Condition Applications in the US or the UK's Gender Pay Gap reports. An AI model weights these sources, adjusts for recency, and outputs a range rather than a single number.

The quality of the estimate depends heavily on sample size. A role like software engineer in San Francisco has tens of thousands of data points, so the range is tight and reliable. A niche role in a smaller city might have fewer than 50 data points, which means the range can be off by 15 to 20 percent in either direction. Always check how many data points the tool is drawing on before you treat the number as authoritative.

• Crowdsourced reports: Glassdoor, Levels.fyi, Blind, LinkedIn Salary
• Posted pay bands: required in California, Colorado, New York, Washington, and several other US states since 2023 onward
• Government filings: US H-1B LCA data, UK Gender Pay Gap Service, Australian WGEA employer reports
• Employer-reported data: some large companies publish pay bands in annual reports or on their careers pages

What inputs produce the most accurate estimate

Garbage in, garbage out applies here. The more specific your inputs, the narrower and more useful the output range will be.

Job title matters less than you might think because titles vary wildly between companies. A senior engineer at one company is a staff engineer at another. Use the level descriptor from the job posting itself, not your current title, and cross-reference it with the responsibilities listed.

• Job title and level: use the exact title from the posting
• Location: city and metro area, not just country; remote roles should use the employer's headquarters state or province for US and Canadian estimates
• Years of total experience and years in the specific function
• Industry: finance and tech pay 20 to 40 percent more than non-profit or education for equivalent roles
• Company size: companies with more than 1,000 employees typically pay 10 to 25 percent above small businesses for the same title
• Equity and bonus: ask the estimator to separate base, bonus target, and equity value so you can compare total compensation

How to read the range you get back

A salary estimator will usually give you a 10th percentile, a median, and a 90th percentile. The median is what most people in that role earn. The 90th percentile is what top earners make, often at well-funded companies with strong negotiation cultures.

Your target in a negotiation should sit between the median and the 75th percentile if you are a strong candidate. If you have a specific skill the posting emphasizes, such as a particular framework or a security clearance, you can justify pushing toward the 90th percentile. Never anchor below the median unless you have a concrete reason, such as a significant gap in required experience.

In the UK, check whether the figure is quoted as annual salary or total package. Pension contributions, private health cover, and share schemes can add 15 to 30 percent on top of base salary at larger employers. In Australia, always clarify whether the number is inclusive or exclusive of superannuation, since the 11.5 percent super guarantee is legally separate from base pay as of 2026.

Turning the estimate into a negotiation script

The estimate is evidence, not a demand. Frame it that way and you will get a better reception from a recruiter or hiring manager.

A simple script that works: 'Based on market data from [source] for this role and location, the range I am seeing is X to Y. Given my background in [specific skill], I am targeting Z. Is that something we can work toward?' This approach names your source, gives a range rather than a single number, and ties the ask to your qualifications rather than your personal needs.

If the employer comes back below your target, ask two questions before deciding. First, is there a review cycle within six months where you could reach your target? Second, is there flexibility in signing bonus, equity, or remote days that closes the gap? A one-time signing bonus does not compound the way a higher base does, so weigh it accordingly.

1. Run the estimate with the exact job title, location, and company size from the posting
2. Identify the 50th and 75th percentile figures as your floor and target
3. Write down two or three specific qualifications that justify the higher end
4. Wait for the employer to name a number first if possible; let them anchor
5. Counter with your target and cite your source by name
6. If they cannot move on base, ask about bonus, equity, and review timeline

Common mistakes that weaken your position

Using a single data point is the most common error. One Glassdoor review or one LinkedIn salary report is not a range. Pull data from at least two sources and note where they agree.

Citing a national average when the role is in a high-cost city is another mistake. The median software engineer salary in the US is not the same as the median in New York City or Seattle. Localize every estimate.

Revealing your current salary before you have an offer weakens your position in most jurisdictions. As of 2026, employers in California, New York, Illinois, and several other US states cannot legally ask for your salary history. In the UK and Australia there is no blanket ban, but you are not obligated to answer. Redirect with: 'I would rather focus on the market rate for this role, which I have researched.'

• Do not use a single data point; cross-reference at least two sources
• Do not use national averages for city-specific roles
• Do not share your current salary before an offer exists
• Do not negotiate against yourself by offering a range with a low floor
• Do not forget to account for cost of living if the role is remote and the employer is in a different city

Using Break-In Bad to prepare before you negotiate

Salary negotiation starts before the offer. It starts with applying to roles where your qualifications genuinely match the level and pay band. Break-In Bad's assistant Omar reads the full job posting, including location, visa requirements, and seniority signals, and scores your fit before you apply. Roles scored 75 percent or above are auto-applied with a tailored resume and cover letter built from facts you confirm, so the application that reaches the recruiter already reflects the experience you will cite in a salary conversation.

When you get to the offer stage, the resume Omar built gives you a clear record of the specific skills and achievements you can point to when justifying a number at the 75th percentile or above.