0.5% startup equity sounds like $250K. Most employees never see a cent

The offer letters are on the screen. One is from a FAANG company: $155K base, $120K in RSUs vesting over four years, a signing bonus. The other is from a Series A startup: $120K base, 0.5% equity. That 0.5% could be worth nothing. It could also be worth more than a decade of FAANG stock grants.

Most engineers accept startup equity with no framework to model what it's actually worth. They see the percentage, feel the upside, and sign. The predictable cash offer from Big Tech looks safer but less exciting. Neither side gets analyzed with the rigor a six-figure decision deserves.

This isn't about salary. It's about wealth-building, risk tolerance, and negotiation leverage in a market that's shifting under everyone's feet.

Why This Matters Right Now

Engineers are making high-stakes compensation decisions amid a volatile IPO market, shrinking startup equity pools, and an AI talent war that has distorted the norms most people will never see. The IPO market in 2025 produced 31 tech deals in the U.S.—more than the prior three years combined, but still a fraction of the 121 deals in 2021. Filings in 2026 are down 11.1% year-over-year. The exit window that startup equity depends on is narrower than it's been in a decade.

At the same time, investors are pressuring startups to reduce large equity pools in favor of capital efficiency. Base salaries have risen 5.8% since 2022, narrowing the gap with Big Tech, but equity grants have reportedly fallen 26% over the same period, according to Carta data cited by Stewart Blake. Companies are restructuring compensation toward cash-heavy packages, positioning equity as upside rather than foundation.

The result: most engineers lack the tools to compare illiquid, preference-affected startup equity against Big Tech's liquid, structured packages. This article provides that framework—funding-stage benchmarks, liquidation math, expected value modeling, and negotiation dynamics—so you can stop guessing and start negotiating with actual leverage.

The Illusion of the "Big Equity" Offer

Startup equity grants look generous on paper. A 0.5% grant at a company valued at $50M sounds like $250K in paper wealth. But the math that matters isn't the grant size—it's what happens when the company exits, and whether there's anything left for common stockholders after investors get paid.

The market has already started correcting for this. Equity grants at startups are reportedly 26% below 2022 levels, while base salaries have climbed 5.8%. Investors want profitability, and that pressure has shrunk equity pools. What used to be a foundational part of startup comp is now framed as upside—a bonus, not a pillar.

Compare that with Big Tech. A FAANG L3 software engineer in 2026 can expect total compensation of $170K–$230K, with $25K–$50K in RSUs vesting predictably over four years, according to Levels.fyi data aggregated on ismysalarygood.com. Google and Meta vest monthly after a one-year cliff. Apple vests quarterly. Amazon uses a back-loaded 5/15/40/40 split. Netflix pays almost entirely in cash salary, with some packages reaching $400K–$900K+. The equity is liquid from day one. You can sell it. You know what it's worth.

A Series A startup offers equity on paper with a four-year vested value ranging from $0 to $1M, according to Startup Search's comparison data. That range is the problem. Without modeling the downside, the number is meaningless.

What Your Startup Equity Is Actually Worth (The Liquidation Math)

Here's what most engineers skip: liquidation preferences. These terms determine whether common stockholders—employees—get anything at all when a company is acquired.

The most common structure in Series A financings is 1x non-participating preferred, used in roughly 70% of Carta-tracked deals as of 2023, according to analysis by Pillar Legal. Under this structure, the investor chooses the greater of their original investment or their pro-rata share of the exit proceeds. It's relatively founder-friendly.

But the landscape is shifting. The proportion of post-Series A financings with senior liquidation preference rose from 29.6% in 2022 to 47.0% in 2023, according to Fenwick data cited on growthequityinterviewguide.com. Deals with multiple liquidation preferences increased to 5.5% in 2023, up from 2.3% in 2021, per the Torys Venture Financing Report 2024. These terms stack the deck against common stockholders.

Walk through a concrete example from Pillar Legal's Series A term sheet analysis. A company raises $1M in Series A at a $10M post-money valuation. It's acquired for $15M.

• 1x non-participating: The investor takes $1.5M (their pro-rata 10% share). Founders and employees split $13.5M.
• 1x participating: The investor takes their $1M back first, then 10% of the remaining $14M. Total: $2.4M. Founders and employees get $12.6M.
• 2x capped participating: The investor takes $1M plus pro-rata share up to a $2M cap. Total: $2M. Founders and employees get $13M.

That's a $900K difference to the common stockholders between the best and worst structures, on a $15M exit. And this is a modest exit.

The Trados case is the cautionary tale cited by Morse Law. In 2005, the translation software company was sold for $60M. Preferred stockholders received $49.2M. Management got 13% through an incentive plan. Common stockholders—employees who had been granted options—received nothing. The preferred stack ate the entire exit.

There is one mitigating factor. The One Big Beautiful Bill expanded the Qualified Small Business Stock exclusion from $10M to $15M for stock acquired after July 4, 2025, as noted in a LinkedIn post by CPA Khaled Alatyieh. If your startup qualifies, you could exclude up to $15M in gains from federal taxes. It's rare, but it's real, and it changes the after-tax math on a successful exit.

Modeling Expected Value—Why FAANG Still Wins on Risk-Adjusted Basis

Paper value is not expected value. To compare startup equity against FAANG RSUs, you need to model probabilities.

Using a standard scenario breakdown for a Series A startup, as outlined by Startup Search:

The expected annual equity value across these scenarios comes to roughly $54,750 per year. Compare that with FAANG L3 RSUs delivering $25K–$50K per year — every year, guaranteed, in liquid stock — on top of a base salary that runs $20K–$45K higher. On a risk-adjusted basis, Big Tech wins for the vast majority of engineers: the certain package matches or beats the volatile one before you even discount for illiquidity.

Even in the best-case startup scenarios—10x to 100x growth, the kind of trajectory Uber saw going from $1B to $76B pre-IPO—the expected value only justifies the risk for specific profiles. AI specialists with 1.5–2x standard equity grants at a pre-IPO company with an S-1 already filed. Engineers who can afford to earn $0 on their equity and still meet their financial obligations. People with genuine conviction about a startup's five-year path, not just excitement about the pitch.

For everyone else, the math favors cash.

The AI Talent War Distorts the Equation—But Only for a Few

Headlines about OpenAI's average stock-based compensation reaching $1.5M per employee in 2025, or Meta offering $100M signing bonuses to lure AI researchers, create a distorted picture of the market. These deals are real, but they're not representative.

Fewer than 1,000 AI research scientists globally can build advanced large language models, according to estimates reported by Invezz and the BBC. That scarcity drives outlier compensation that doesn't extend to the broader engineering workforce. The median startup engineer in 2026 earns a $120K base with 0.2%–0.75% equity, per FounderMath benchmarks. Nowhere near AI-tier packages.

The Thinking Machines Lab situation illustrates the volatility. Five founding members of the startup—founded by former OpenAI CTO Mira Murati—left for Meta despite the company raising $2B at a $12B valuation, according to Invezz. Even massive funding and high-profile leadership couldn't retain talent when Big Tech waved enough cash.

Meanwhile, the talent acquisition playbook has gotten unconventional. Microsoft hired Inflection AI's co-founders and team via a $650M deal. Amazon licensed Adept's technology and brought in key team members. Google paid roughly $2.4B to bring in Windsurf co-founder Varun Mohan. These are acqui-hire structures, not standard compensation packages, and they signal how aggressively Big Tech will move to secure scarce AI talent.

If you're not one of those roughly 1,000 people, don't let headline comp distort your personal risk calculus.

When Startup Equity Does Make Sense

Startup equity is rational only under narrow conditions. Here are the cases where it genuinely works:

Pre-IPO with a clear timeline. If the company has filed an S-1 or has a credible IPO plan within 12–18 months, the liquidity horizon is defined. You're not betting on a five-year gamble—you're betting on a near-term event with public documentation.

AI/ML specialized roles. AI and ML engineers at startups reportedly command 1.5–2x standard equity grants due to talent scarcity, according to Fonzi's equity compensation guide. If you're in this category, the expected value calculation shifts in your favor, especially at the seed or Series A stage where grants range from 0.5%–1.5%.

Early-stage conviction bets. If you deeply understand the startup's market, believe in the founding team, and the expected value aligns with your personal risk tolerance, an early-stage bet can make sense. Seed-stage grants of 0.5%–1.5% at a company you genuinely believe in are a different bet than 0.2% at a Series A company you found on LinkedIn last week.

These are exceptions. The IPO market is still weak. Most startups fail. Most equity grants expire worthless. Go in with open eyes.

Negotiation Leverage Is on Your Side—Use It

Candidates have more power to push for cash than they've had in years. Companies expect pushback on base salary. Employee preference for certainty is rising after years of down rounds and delayed IPOs, according to Mondo's 2026 compensation analysis cited by Stewart Blake. Smaller equity grants mean less lock-in, which means more mobility for you and more pressure on employers to compete on cash.

Use a FAANG offer as leverage to negotiate a higher base at a startup. If the startup can't match the cash, ask for protective terms: acceleration on change of control, extended exercise windows after departure, or a larger equity grant that compensates for the cash gap.

Base-pay increases for U.S. tech workers are projected at 3.5% for 2026, down from 4% in 2025, per Bluesignal data. The window for locking in strong cash comp is narrowing. If you have an offer in hand, the time to negotiate is now—not after you've signed and the leverage is gone.

A Decision Framework You Can Use Today

Evaluate any offer using three factors: liquidity, probability, and personal alignment.

Step 1: Compare liquid cash over four years. Add base salary and any guaranteed or near-guaranteed equity (vested RSUs at public companies). This is your floor.

Step 2: Model equity value using stage-specific benchmarks and liquidation math. For startups, apply the probabilistic scenarios above. For public companies, use current stock price and vesting schedule. Don't use the startup's paper valuation as the equity number—use the expected value after liquidation preferences.

Step 3: Assess your risk horizon. Can you afford $0 on your equity and still pay rent? Do you believe in this startup's five-year path enough to bet on it? Are you optimizing for maximum upside or financial stability?

Here's a simplified comparison for an entry-level software engineer in 2026, drawing on data from ismysalarygood.com and Startup Search:

The FAANG offer delivers comparable expected value with full liquidity and far lower risk. The startup offer delivers a wider range—from zero to life-changing—but once you discount for illiquidity and variance, the weighted average favors cash.

Know Your Walk-Away

The best offer isn't the biggest number on paper. It's the one whose risk you understand, whose terms you've modeled, and whose walk-away point you defined before you opened the laptop.

Don't bet your career on a spreadsheet fantasy. Model the math, know your walk-away, and sign with clarity.

Working in frontier tech? Zero G Talent tracks the openings: browse frontier tech jobs, the companies hiring, and the people building the field.