In 2019, researchers at Stanford published a study on salary transparency that produced a finding employers have quietly hoped you'd never read: when workers have access to accurate market-rate data, they are paid 4–8% more than workers who don't. The gap isn't talent. It's information asymmetry. The company knows what the role is worth on the open market. In most cases, you're negotiating blind.
That changes the moment you treat your compensation as a data problem rather than a personal one.
Market rate is not how much you feel you deserve. It's the price a specific type of employer in a specific location is currently willing to pay for a specific configuration of skills. Strip the emotion out of it. Your job in this lesson is to build a credible data structure - a range with a floor, a target, and a ceiling - that you can defend with sources, not just confidence.
Why one salary site is never enough
Glassdoor reports averages. Averages are arithmetic fiction. They blend the junior developer at a 15-person startup with the senior one at a Fortune 100. When you look at a number like "$112,000 median for Product Manager," you're looking at a smoothed curve that obscures everything useful.
To triangulate a real range, you need three types of sources operating simultaneously.
Public aggregators like Glassdoor, Levels.fyi, and Payscale give you baseline orientation. They're the starting point, not the destination. Filter aggressively: job title, company size, industry, and geography. A software engineer's salary in Austin and in San Francisco aren't the same data point regardless of what any single number suggests.
Industry-specific surveys are a level above this. Professional associations - the Project Management Institute, the Society for Human Resource Management, the IEEE - publish annual compensation surveys that are narrower, peer-verified, and segmented by seniority in a way that public sites can't match. These are often paywalled, but your professional association membership fee is significantly cheaper than a missed negotiation.
The highest-quality real-time signal comes from recruiters in your specialty. A recruiter who places five people a month in roles like yours knows exactly what offers are closing right now - not what they closed 18 months ago when the dataset you're reading was compiled. A 15-minute call framed as an informational conversation will return better compensation intelligence than hours on job boards.
Key Point: Always filter salary data to the last 6–12 months and to your specific industry segment. Compensation in 2023 is not compensation in 2025. Using stale data in a negotiation signals to an informed employer that you haven't done the work.
The variables that move your number
Three factors account for most of the variance in compensation for equivalent roles.
Role specificity is the first. Job titles are marketing, not specifications. A "Senior Analyst" at a regional insurance firm and a "Senior Analyst" at a global asset manager share a title and little else. When you're gathering data, look at job descriptions for currently open roles - specifically the responsibilities section and the "preferred qualifications" list. Match yourself to the functional reality, not the label.
Geography still matters, even post-pandemic. Remote roles have compressed differentials somewhat, but companies headquartered in high-cost labor markets still pay premiums because they're competing with other employers in those markets. When negotiating a remote role with a San Francisco headquarters, that geography anchors the budget even if you're working from elsewhere.
Specialized scarcity is the variable most people underweight. If you hold a certification, a clearance, or deep expertise in a tool or methodology that fewer than 5% of candidates in your category have, you are not competing on the same supply curve as everyone else. Your floor moves up simply because the pool of qualified substitutes is smaller. Identify your scarcity factor before you sit down to negotiate.
Building your target range
Don't negotiate to a single number. Build a range with three reference points.
Your floor is the minimum you'd accept given the full compensation picture - base, bonus, equity, and benefits converted to dollar equivalents. It's grounded in your data, not your anxiety about what the employer might say.
Your target is the realistic number supported by your data and your specific differentiators. This is not a stretch - it's what the market actually pays for someone with your profile at companies comparable to this one.
Your ceiling is the high end of the verified range, defensible with data, used as your opening number. The anchor effect in negotiation is well-documented: the first number spoken sets the psychological gravity of the conversation. Starting high - with data - pulls the final settlement upward even when the employer negotiates you down.
The range also gives you options mid-conversation. If the employer can't match base, you have room to pivot to sign-on bonuses, equity, or benefits without losing the thread.