The Same Job Title Spans More Than 2x in Pay Across Employers

Median of the midpoint of each advertised salary range, for the 29 employers with 40 or more priced engineering postings on a single ATS. Holding the board and the role family constant removes the two composition effects that usually explain gaps like this. What remains is the employer.

Postings whose title contains Software Engineer and carries neither a seniority word nor a numeric or roman level. That second exclusion matters: employers who encode level as a number, such as Software Engineer 5 or Software Engineer Level 2, are advertising a seniority the words do not show, and leaving them in inflated this range at both ends. Google is highlighted as the baseline, being the only employer here with a sample large enough to stand on its own. Read this table as illustration; the headline figure comes from the 2,479 posting cohort in the chart above.

Source: Seeker job corpus, 288,411 active job postings, snapshot 2026-08-25. seekerscore.com/insights/same-title-different-pay-2026

The comparison

Take every engineering posting on a single applicant tracking system that publishes a US dollar salary range, and keep the 29 employers contributing at least 40 of them. That is 2,479 postings. Rocket Lab sits at a $143,550 median. xAI sits at $310,000. Between them, in order, are Databricks at $216,104, Coinbase at $202,482, Datadog at $174,000 and MongoDB at $165,000. These are not obscure differences between industries. They are companies hiring the same kinds of engineers, advertising on the same platform, in the same months.

The control that makes it interesting

A pay gap between employers is not surprising on its own. What makes this one worth reporting is what happens when you measure the variable everyone assumes is dominant, inside the exact same cohort. Postings asking for 1 to 2 years of experience carry a median of $137,950. Then $175,500 at 3 to 4 years. Then $200,500 at 5 to 7. Then $227,000 at 8 to 10. That is a clean, monotonic gradient with no reversals, and it is worth 1.65x end to end. The employer gradient in the same data is worth 2.16x. Nine years of accumulated experience moves the number less than the choice of who you send the application to. It is worth adding that the employers at either end are not asking for different things. The median posting at xAI asks for 5 years. The median posting at Rocket Lab asks for 5 years.

What this means if you are searching

Searching by job title is the default behaviour on every job board, and this is the cost of it. A title tells you roughly what the work is. It tells you very little about what the work pays, because the widest variable is not encoded in the title at all. The practical version: if you are weighing whether to spend six months acquiring the experience that moves you from one band to the next, the same six months spent widening the set of employers you apply to has a comparable or larger expected effect on the offer. This is not an argument against building experience. It is an argument against treating employer selection as the part of the search you do not think about.

Methodology

Population is active postings in the Seeker corpus with a salary range denominated in US dollars, an annualized figure of at least $10,000, and an upper bound below $2,000,000. Those filters exist because a minority of postings carry hourly rates or non US currencies, and mixing them produces nonsense. Salary is the midpoint of the advertised range. All figures are medians, not averages. The employer cohort is restricted to a single applicant tracking system and to the engineering role family, which holds constant the two composition effects that most often explain gaps of this size: which board an employer publishes on, and which mix of roles they are hiring for. Each employer contributes at least 40 postings. The illustration table applies one further exclusion: titles carrying a numeric or roman level, because an employer writing Software Engineer 5 has stated a seniority even though no seniority word appears.

What this does not show

These are advertised ranges, not offers, and not total compensation. Equity is absent from every figure here, which matters most at exactly the companies sitting at the top of the chart. The roles inside each employer are similarly titled rather than identical, so some of the employer gap is a difference in the seniority mix that titles do not capture, even though median stated experience sits at 5 to 7 years for nearly every employer in the cohort. Roughly a third of active postings publish a salary at all, and employers that disclose are not a random sample of employers. The cohort is one ATS, which skews toward venture backed technology companies and is not a picture of the whole labour market. Finally, this is a snapshot of what is being advertised now, not a trend.

See where you actually fit

If the employer axis is the one carrying the variance, the useful question is not which title to search for. It is which employers are hiring for work you can already do. Upload a resume and Seeker will map your background against live postings, including the ones whose titles you would never have searched.

Frequently asked questions

Related Data

Guides

• Salary Negotiation →
• Job Titles Explained →
• What Jobs Am I Qualified For? →