The Promotion vs. Job-Hop Math: Why Staying Loyal Costs Senior Engineers $300K

There is a story engineers tell themselves when they decide to stay at a company for a fifth or sixth year. It is a story about stability, about vesting cliffs, about not wanting to be seen as a job hopper. It is a story about loyalty being rewarded, about the next promotion cycle being the one that finally resets the comp to where it should be. It is a comfortable story. It is also, on average, a very expensive one.

The data on the promotion-versus-job-hop question has been accumulating for years, and the verdict is not particularly close. Senior engineers who stay at the same company average 3 to 4 percent annual raises.^[1] Those who move to a new employer routinely collect compensation increases of 15 to 30 percent — sometimes more.^[2] Over a ten-year career, the cumulative gap frequently exceeds $300,000 in total compensation, and in high-cost markets like San Francisco or New York, it can exceed $500,000.

This is not an argument that you should always job-hop, or that loyalty has no value whatsoever. It is an argument that you should make the decision with accurate information instead of comforting myths. Most engineers dramatically underestimate the financial cost of staying put. This article runs the math and explains the structural reasons why the gap exists — so you can decide what your loyalty is actually worth to you.

The Loyalty Math Nobody Talks About

The reason the loyalty penalty exists is structural, not a matter of bad luck or bad managers. Companies budget for annual raises as a fixed percentage of payroll, and that percentage has hovered between 3 and 4 percent for most of the last decade, with modest exceptions during the pandemic-era talent wars of 2021 and 2022.^[3]

Now compare that to what happens when you interview externally. The Federal Reserve Bank of Atlanta's Wage Growth Tracker, which separates wage growth for job-switchers from job-stayers, consistently shows a gap of 5 to 8 percentage points in favor of those who change employers.^[5] That is before accounting for the fact that engineers negotiate their offers — and engineers who receive competing offers negotiate them upward by an average of 11 percent according to Payscale's Compensation Best Practices Report.^[6]

The arithmetic is brutal once you run it forward. An engineer earning $180,000 who receives a 3.5% raise annually will earn $214,000 after five years. An engineer who starts at the same $180,000 but switches jobs twice in five years — each time for a 20% increase — will be earning $259,000 by year five. The difference in annual compensation at year five alone is $45,000. Summed across all five years, the switcher has collected roughly $115,000 more in total compensation than the stayer, even accounting for a few months of job-search friction.

How Raises Work (and Don't Work) Inside Companies

To understand why internal compensation growth is structurally limited, you need to understand how companies actually budget and allocate raises — a process that is deliberately obscured from most employees.

Almost every company with more than 200 employees operates on a merit increase budget: a fixed percentage of total payroll that HR allocates to managers for distribution across their teams.^[7] WorldatWork's annual salary budget surveys have tracked this number for decades. The 2025 survey found that the median merit increase budget across industries was 3.5%, down from the 4.2% peak in 2022.^[8]

Here is what that budget means in practice. Your manager has a pool of money equal to 3.5% of their team's total salaries. If they want to give you a 6% raise — which might be closer to what your market value increase actually warrants — they have to give someone else on the team a 1% raise, or nothing at all. Most managers are not willing to have that fight. So raises get compressed toward the average, and the distribution clusters tightly around 3 to 4 percent regardless of actual performance.^[9]

Promotions change the math, but not by as much as most engineers believe. A typical promotion from senior engineer to staff engineer carries a salary increase of 10 to 15 percent at the moment of promotion.^[10] But promotions at the senior and staff level happen on average every three to four years — and they require sustained exceptional performance, organizational advocacy from multiple levels of management, and often a specific business case that you have little control over. The expected value of waiting for a promotion, when discounted by the probability that it actually happens on schedule, is frequently lower than the certain 20% increase available in the external market right now.

The Real Numbers: Job-Hopping vs. Staying

The compensation data on job-switching has become more transparent over the last five years, largely because engineers started sharing their offer letters and salary histories on platforms like Levels.fyi and Glassdoor.^[11] The picture that emerges is consistent and sobering for loyal employees.

The ADP Research Institute's Workforce Vitality Report, which tracks wage data for millions of American workers, found that job-switchers consistently out-earn job-stayers by a margin that has widened since 2020.^[12] In the technology sector specifically, the gap between switcher and stayer wage growth peaked at 9 percentage points during the 2021 talent wars and has since settled around 6 to 7 points.^[13]

Even more revealing is the data on what happens to compensation relative to market rate over time. Mercer's annual survey of technology compensation found that engineers who stay at a single employer for more than four years are compensated at an average of 88% of current market rate for their role and level — meaning they are, on average, leaving 12% of their potential earnings on the table simply by not testing the market.^[14] After six years, that figure drops to 81 percent of market.

Why Comp Bands Trap Senior Engineers

Compensation bands are the structural mechanism that caps how much any company can pay someone at a given level, and they are one of the most important — and least discussed — forces in engineering compensation. Every level at a well-structured company (L5, L6, L7 in Google's system; E5, E6 at Meta; Senior, Staff, Principal at many others) has an associated compensation band: a minimum, midpoint, and maximum salary range.^[15]

The problem for loyal senior engineers is simple: once you have been in a role for several years, you tend to be at or near the top of your band. And the top of the band is not the ceiling in theory — it is the ceiling in practice. Companies rarely pay above the maximum of a band without either promoting the person or reclassifying the role. When you are at the top of the Senior Engineer band and not yet ready for a Staff promotion, you are, by definition, receiving raises against a number that is already close to its structural ceiling.

New hires do not have this problem. When a company recruits an experienced engineer from outside, they benchmark the offer against current market data — which reflects the actual competitive rate — not against the internal band structure. A new hire at the same level and experience as a three-year veteran frequently earns 15 to 25 percent more than the veteran, a phenomenon that compensation researchers call "internal pay compression."^[16]

The irony is sharp: the company's most experienced, most embedded engineers — the ones who have spent years building institutional knowledge and mentoring junior staff — are frequently the most underpaid relative to market. The engineers who just arrived last month, with no context and no relationships, are being paid what the market actually demands. Harvard Business Review has documented this dynamic extensively, noting that companies routinely pay 20 to 30% more to recruit externally than they would need to spend to retain internally — but they structure their raise budgets in ways that make external poaching inevitable.^[17]

The $300K Gap: A 10-Year Case Study

Abstract percentages can obscure the scale of what is at stake. Running the actual numbers across a concrete ten-year career trajectory makes the magnitude of the decision real.

The cumulative gap in this scenario: approximately $330,000 in total compensation over ten years. The gap is not just in base salary — RSU refreshes at new employers frequently reset the equity clock at current valuations, which in growing companies can be substantially more valuable than vesting into shares granted years earlier at a lower strike price.^[18]

It is worth noting what Engineer B's path did not require: burning bridges, abandoning vested equity, or moving into roles they were unqualified for. Each move was a step from a position of strength, with a legitimate competing offer that validated the market value at that moment. The timeline — moves at years 2, 5, and 8 — is conservative by the standards of the modern tech labor market, where the median job tenure for software engineers is approximately 2.5 years according to LinkedIn's Economic Graph data.^[19]

The Career Capital You Are Leaving on the Table

Compensation is the most quantifiable dimension of the loyalty penalty, but it is not the only one. Career capital — the combination of skills, relationships, and opportunities that compound over time — also accretes differently depending on whether you move or stay.

Engineers who move across companies are systematically exposed to a wider range of technical problems, organizational structures, and engineering cultures. A senior engineer who has worked at a large-scale consumer platform, a growth-stage startup, and an enterprise software company in their first eight years of career has a depth of contextual knowledge that is extremely difficult to develop within a single organization. MIT Sloan Management Review research on career mobility found that professionals with multi-company experience are 40% more likely to reach senior leadership positions than those who have spent their careers at a single firm, controlling for talent and role level.^[20]

There is also the matter of network breadth. Your professional network at a single company, no matter how large the company, is a structurally limited graph — it is dense within the organization and sparse outside it. Each company you work at adds an entirely new cluster of nodes: former colleagues who go on to found startups, join promising teams, or end up in positions to refer you to roles that are never publicly posted. The SHRM research on hiring channels consistently shows that 70 to 80 percent of roles are filled through networks rather than public applications.^[21] An engineer with a broad, multi-company network has structural access to that hidden job market that a deeply loyal single-employer engineer simply does not.

Title and level progression also tends to move faster externally. At large companies, promotions from Senior to Staff to Principal require extensive evidence packets, multiple calibration cycles, and organizational alignment across skip levels. At smaller or faster-growing companies, the level inflation that happens during rapid hiring frequently allows engineers to enter at a higher level than they currently hold — a phenomenon that Korn Ferry's talent research describes as "level arbitrage," where engineers gain one to two levels of seniority by moving to a company whose hierarchy is less compressed.^[22]

When Loyalty Actually Pays (And When It Doesn't)

The case for job-hopping is not absolute. There are specific situations where staying at a company is the financially and professionally rational decision, and conflating those situations with the general pattern leads to bad choices in either direction.

Equity with real upside. If you are at a pre-IPO company or a company in an early-stage hypergrowth phase where the equity has genuine, non-speculative upside, the expected value of vesting that equity can dwarf any compensation increase available in the market. The calculus changes entirely when your unvested RSUs could be worth $1 million in 18 months. This is a legitimate reason to stay — but it requires an honest, unsentimental assessment of the equity's actual likelihood of vesting and actual probability of being worth what you think it is. Most unvested equity at private companies is worth significantly less than the paper valuation implies.^[23]

A specific, time-bound opportunity. If you are three months away from leading a launch that will meaningfully change your resume — the kind of work that demonstrates you can operate at the next level — that is a legitimate reason to delay a move. The key word is time-bound. "There will be better opportunities here next year" is not a specific opportunity; it is a deferral strategy that frequently repeats itself indefinitely.

Active sponsor relationships. Research on career advancement, including Herminia Ibarra's work at London Business School and Catalyst's sponsorship research, consistently shows that having a senior leader actively advocating for your advancement is worth more than almost any other career variable.^[24] If you have a sponsor — not a mentor, but someone with organizational power who is actively putting your name in rooms you are not in — that relationship is rare and valuable, and abandoning it for a modest comp increase may be a poor trade.

The situations where loyalty reliably does not pay, despite frequent claims to the contrary: a vague promise of promotion "in the next cycle"; a manager who consistently says you are "on track" but whose advocacy has not been tested; a company that has recently conducted layoffs and is using loyalty as social pressure to discourage exit interviews; and any situation where you have been in a role for more than four years without either a meaningful promotion or evidence that your compensation has kept pace with market.

How to Job-Hop Without Burning Bridges or Your Career

The term "job-hopper" carries a stigma that, in the engineering labor market, has largely been revised by data. Indeed's Hiring Lab research found that hiring manager bias against candidates with multiple short tenures has declined sharply since 2019, with the majority of technical hiring managers now considering two-year tenures standard and three moves in ten years unremarkable.^[25] That said, how you move matters almost as much as whether you move.

Leave with a complete cliff vest and a positive relationship. The minimum viable tenure at any company is long enough to vest your first cliff — typically one year. Leaving before that date is almost always a financial mistake, and it is the kind of move that hiring managers do notice. Staying through the first cliff, building a track record of shipped work, and leaving on terms that allow you to use your manager as a reference takes almost nothing away from the compensation strategy and adds significant career protection.

Do not leave reactively. The engineers who burn bridges are not those who move every two to three years with a plan — they are those who leave abruptly after a conflict, a denied promotion, or a bad review cycle. Reactive departures give you less negotiating leverage, shorter runways to find the right role, and the kind of bitterness that comes through in reference checks. Move from strength, not frustration. Start your search while you still feel good about your current role.

Interview before you are desperate. The best time to interview is when you do not need to move. Candidates who are comfortable in their current role negotiate better, accept offers more selectively, and are perceived as more attractive by hiring teams — because they are perceived as recruited, not job-hunting. National Bureau of Economic Research labor market studies on offer negotiation found that employed candidates receive initial offers that are, on average, 8 to 12 percent higher than those extended to unemployed candidates with identical credentials.^[26]

Use competing offers as a correction mechanism, not a bluff. A real competing offer — one you would genuinely accept — is the single most effective compensation correction tool available to an engineer at any company. Companies that would not give you a 10% raise in the annual review cycle will frequently match or exceed a competing offer, because the cost of recruiting and onboarding your replacement is estimated at 50 to 200% of your annual salary.^[27] If you get a counter-offer, evaluate it carefully: our analysis of counter-offer dynamics walks through when accepting one makes sense and when it is a delayed exit anyway.

Prepare rigorously for each move. The financial returns from job-hopping are concentrated in the negotiation, and the negotiation is determined by how well you perform in the interview process. An engineer who gets through five rounds and then fails the system design interview forfeits the entire compensation advantage of moving. The preparation required for a senior or staff-level move — system design depth, behavioral interview clarity, and crisp articulation of past impact — is substantial and is worth investing in seriously. Our guide to salary negotiation strategies covers the mechanics of anchoring and counter-offering once you have an offer in hand.

The decision to stay or move is not a moral one. Companies are not loyal to their employees in the structural sense — they lay people off when market conditions change, restructure teams when priorities shift, and hire externally at market rate while constraining internal raises to budget percentages. You are not obligated to subsidize their retention by accepting below-market compensation. Treating your career as a business — one that benchmarks its price periodically against what the market will actually pay — is not cynicism. It is financial literacy.

The engineers who collect the most career capital and the highest lifetime compensation are not the ones who hop recklessly or who stay out of inertia. They are the ones who understand the mechanism well enough to make deliberate, timed decisions: staying when there is a genuine, specific reason to stay, and moving when the market has diverged from what they are being paid. Run the math on your own situation at least once a year. The numbers may surprise you.