RSU Vesting Schedule: Google Front-Load vs Amazon Back-Load – Which Pays You Faster?

The candidate sitting across from me in the Google L6 offer negotiation in March 2022 thought she’d done the math. She hadn’t. She’d compared the two-year total at Google against Amazon’s package, missed the vesting asymmetry entirely, and walked away convinced Amazon paid more. She was wrong by $340,000 in realized income over forty-eight months.

RSU vesting schedules are not fine print. They are the entire economic structure of your employment. Front-load versus back-load is not a preference. It is a capital event timing decision that alters your risk profile, your tax planning, and your bargaining position in ways most candidates never examine until they have already signed.

What Does “Front-Loaded” Actually Mean in Google’s RSU Structure?

Google’s vesting is aggressively front-loaded. A typical L5 offer in 2023: 25% vests at the one-year cliff, then quarterly thereafter. But here’s what the recruiter does not emphasize.

Google also runs a “refresh-heavy” compensation model where the initial grant is smaller than the projected four-year total because they expect substantial additional grants—“refreshes”—to stack on top. The front-loaded structure is designed to make those refreshes meaningful faster. A Google L6 PM I debriefed with in Q2 2023 had a $280,000 initial grant vesting 25/25/25/25, but by month eighteen had accumulated $410,000 in unvested refreshers on top. The front-load makes the refreshes hit earlier in their own vesting cycle.

This creates a compounding effect that candidates from flat-vest companies do not intuitively grasp. Amazon’s back-loaded structure—5/15/40/40 or the newer 5/15/20/20/20/20/20 for six-year grants—deliberately suppresses year-one and year-two realizable value. The Amazon L7 I sat in an HC for in 2022 had a $380,000 grant that paid out $19,000 in year one and $57,000 in year two. Same role at Google L6: $310,000 grant, $77,500 year one. The Amazon candidate needed the stock to appreciate 40% just to break even on realized value against the Google package in year two.

The psychology here is deliberate. Amazon’s back-load enforces retention through economic hostage-taking. Google’s front-load accelerates your portfolio concentration in GOOGL, which then psychologically binds you to the refresh cycle. Neither is your friend. Both are engineered.

How Does Amazon’s Back-Loaded Vesting Actually Work in Practice?

Amazon shifted from the infamous 5/15/40/40 to a six-year structure in 2022: 5/15/20/20/20/20. The recruiter in my September 2022 debrief for the Alexa Shopping team described this as “more competitive with the market.” It was not. It was longer. The back-load got worse in duration, not better in structure.

Here is the specific scenario from that debrief. Candidate had a $400,000 grant at L6. Year one: $20,000. Year two: $60,000. Year three: $80,000. Year four: $80,000. Year five: $80,000.

Year six: $80,000. Total nominal value: $400,000. But the candidate’s alternative was a Google L6 offer at $340,000 front-loaded. At 15% annual stock appreciation—a conservative estimate for both companies during 2022-2024—the Amazon package’s realized value at the end of year four was $353,000 in stock received, versus $340,000 at Google. Break-even. Except the Google candidate had $85,000 in year one to invest, $85,000 in year two. The Amazon candidate had $20,000 and $60,000. Time value of money at 5% risk-free rate made the Google package worth $23,000 more in net present value even before considering the refresh differential.

The hidden mechanism: Amazon’s back-load coincides with their “regret minimization” retention strategy. Employees in years three and four face a cliff of unvested value that makes leaving extraordinarily expensive. In a 2023 AWS debrief for an L7 Solutions Architect, the hiring manager explicitly noted: “We need to see four-year commitment probability.” The vesting schedule is the enforcement mechanism. The problem is not the total value. It is the optionality you surrender.

Which Vesting Structure Pays More in the First Two Years?

Google front-load pays more in cash-equivalent realized income in years one and two. Full stop. The specific numbers from a 2023 offer comparison I reviewed: Google L5, $265,000 base, $320,000 RSU grant, 25/25/25/25 vesting. Amazon L6, $160,000 base, $380,000 RSU grant, 5/15/20/20/20/20 vesting. Year-one realized RSU: Google $80,000; Amazon $19,000. Even with Amazon’s higher base in the L6 band, the total first-year compensation gap was $61,000 in Google’s favor. The candidate chose Amazon for “total package” and regretted it at month fourteen when his mortgage rate reset.

The counter-intuitive insight: total nominal grant value is a decoy metric. In a 2023 Google Cloud HC for an L6 Product Manager, the committee explicitly discussed “annualized realized comp” versus “grant face value.” The candidate had a competing Amazon offer with a $450,000 nominal grant versus Google’s $380,000. The committee approved a $40,000 base increase and a $50,000 sign-on bonus specifically to highlight the front-loaded realized advantage. The candidate still chose Amazon.

He left Amazon in month twenty-two, forfeiting $287,000 in unvested back-loaded RSUs. I saw his LinkedIn. He did not make a mistake. He made a rational choice with incomplete information.

The “not X, but Y” here: the question is not which company has the larger number on the offer letter. It is which structure aligns with your liquidity needs, your risk tolerance, and your probability of staying four-plus years. Google front-load rewards the mercenary who captures value early. Amazon back-load rewards the institution-builder who stays. Neither structure cares about your personal financial situation.

How Do Refresher Grants Change the Vesting Math Over Time?

Refresher grants are where Google front-load compounds into a genuinely different wealth trajectory. Amazon does grant refreshers, but the structure blunts their impact. At Google, a refresh granted in year two begins vesting on its own front-loaded schedule immediately, stacking atop the original grant’s ongoing vesting. At Amazon, a refresh granted in year two begins vesting on the same back-loaded schedule, meaning its meaningful value does not arrive until years three and four of its own cycle—if you are still there.

Specific scenario from a 2024 debrief: Google L7 engineer, joined 2019, initial grant fully vested. By 2023, annual refresher accumulation meant $680,000 in unvested RSUs, all on overlapping front-loaded schedules. She was effectively “stuck” in a high-positive way—leaving meant forfeiting two years of front-loaded refreshes that would pay out immediately.

Amazon L7 counterpart from same cohort: $920,000 in unvested RSUs, but on back-loaded schedules where 60% of value sat in years five and six of each grant’s life. His effective “walk-away cost” was higher in nominal terms but lower in near-term realizability. The Google structure creates more frequent, more liquid wealth events. The Amazon structure creates longer-dated, more illiquid retention hooks.

The framework Google uses internally is “total unvested value at 12-month intervals.” Recruiters are trained to project this number. In a 2022 compensation training for hiring managers, the Google People Ops lead explicitly instructed: “Show the candidate their unvested total at month 24, not month 48. The 24-month number is what drives acceptance.” This is not generosity. It is presentation engineering. The front-load makes the 24-month number look enormous because it includes two full years of refresh initiation.

What Happens to Unvested RSUs If You Leave Before the Cliff?

At Google, cliff vesting means zero before year one, then quarterly. At Amazon, the 5% year-one vest means you leave with almost nothing if you depart at month eleven. The asymmetry is brutal and deliberate.

I sat in an Amazon-device HC in 2021 where a candidate’s previous employer was Amazon, and she had left at month ten of her second year. She forfeited $94,000 in unvested value. The Google hiring manager’s comment in the debrief: “She knows the cost of leaving. She won’t make that decision lightly here either.” The front-load was specifically cited as a retention advantage.

The specific policy detail most candidates miss: Amazon’s six-year vesting includes no acceleration on acquisition, no proration on termination without cause in most cases, and explicitly excludes vesting during any performance improvement plan duration. Google’s front-loaded structure at least concentrates your at-risk period earlier. If you survive year one at Google, 25% is secure.

At Amazon, year one is 5%—barely worth the tax complexity of filing for it. The “not X, but Y” is stark: the risk is not leaving money on the table. It is structuring your life around unvested wealth you do not control.

A candidate in a 2023 Google Search debrief asked about “cliff protection.” The hiring manager’s response, verbatim: “We don’t do pro-rata before cliff. No one here does. The year-one vest is the commitment test.” This applies at both companies. But Google front-load makes passing that test economically meaningful. Amazon’s 5% is a rounding error.

Preparation Checklist

• Model your personal liquidity needs month-by-month, not year-by-year, before comparing offers. The PM Interview Playbook has a compensation calculator with real vesting scenarios from Google and Amazon debriefs.

• Request specific vesting schedules in writing, including refresh grant policy documents, not recruiter verbal summaries.

• Calculate net present value at 5% and 10% discount rates for any competing offers.

• Ask directly: “What percentage of L6 hires at this level receive refresher grants in year two, and at what median dollar value?”

• Verify cliff terms: acceleration, proration, and performance-plan exclusions.

• Build a departure-cost model: unvested value at month 12, month 24, month 36 for each offer.

• Compare base salary plus first-year realized RSU, not total grant value, for immediate liquidity planning.

Mistakes to Avoid

BAD: Comparing $400,000 Amazon grant to $320,000 Google grant and concluding Amazon pays more.

GOOD: Modeling Amazon year-one realization at $20,000 versus Google at $80,000, then adding base, sign-on, and refresh probability.

BAD: Assuming “I’ll stay four years anyway, so vesting structure doesn’t matter.”

GOOD: Recognizing that 40% of Amazon L6 hires in the 2022 class did not reach year four, per internal attrition data discussed in a 2024 hiring manager roundtable, and pricing that optionality loss.

BAD: Negotiating for “higher total comp” without specifying vesting schedule changes.

GOOD: At Google, negotiating for an $80,000 sign-on bonus to bridge the year-one gap if front-load is already maximized; at Amazon, explicitly requesting first-year vesting acceleration (rarely granted, but the ask signals sophistication).

FAQ

Does Google front-load apply at all levels, or only L6 and above?

Front-load applies universally, but the refresh multiplier dominates at L6+. At L4 and L5, the base grant is modest enough that refreshers do not accumulate dramatically. A 2023 Google Finance HC for an L5 role showed median refresh at 40% of initial grant value, versus 80-120% at L6. The front-load structure exists at all levels, but the wealth compounding effect is not meaningful until L6.

Can I negotiate vesting schedule changes at either company?

Rarely at Google, occasionally at Amazon for executive levels. In a 2022 Amazon Retail L8 offer, the candidate negotiated six-year vesting down to four-year with 10/20/30/40 structure. Cost to Amazon: $45,000 in additional year-two compensation. Gain to candidate: ability to realize value fast enough to cover a second child’s tuition. At Google, vesting schedule is typically non-negotiable below L8; negotiate sign-on and base instead.

How do I compare these structures if I plan to stay less than two years?

You should not join Amazon with a sub-two-year horizon. The 5% year-one vest plus 15% year-two means you capture 20% of nominal grant value in twenty-four months. Google front-load gives you 50% in the same period. In a 2023 debrief for a candidate leaving Meta for Google with explicit two-year plans, the hiring manager approved a $120,000 sign-on specifically because “we know she’s not staying for refreshes, make the front-load work harder.” Mercenary candor, rewarded.

---

Ready to build a real interview prep system?

Get the full PM Interview Prep System →

The book is also available on Amazon Kindle.