Google's Stock Couldn't Retain Talent, So Facebook Changed the Playbook?

Deep Thoughts on AI and Aspirations —— ByteDance Thought Circle

When Google went public in 2004, it minted a generation of young millionaires. Side effects emerged quickly: once their initial stock options vested, some engineers cashed out and left for Facebook, then an unimpressive startup desperately hunting for talent. Silicon Valley called this “graduating”—graduating from Google.

Facebook benefited from this talent migration while recognizing something crucial: when they went public, the same thing would happen to them, only more intensely. Employees holding years of accumulated RSUs with virtually zero cost basis would trigger what could be the largest arbitrage and talent exodus in internet history once those shares became liquid.

Connect these two events and the real problem with equity compensation surfaces. It was never about “how much to distribute”—it’s about “how to structure vesting over time.” Most companies obsess over total amounts and percentages when discussing incentives; a few design temporal structures. These two companies’ approaches showcase both sides of this craft.

What Retains People Isn’t the Stock, It’s the Unvested Portion

Start with a basic concept. Granting and vesting are two different things: granting is a promise, vesting is delivery. The time gap between promise and delivery is where incentives actually work. Stock is just the vehicle; the time gap is the product.

Google’s lesson lies here. Silicon Valley convention allowed departing employees to keep vested options, so once those large initial grants vested completely, the retention rope snapped. With exercise prices near zero and stock prices up dozens of times, people left with paper wealth while the company retained nothing.

Facebook’s later executive design went the opposite direction. Incentive shares were granted annually, but previously granted portions had to be mostly vested before new grants began vesting, then vested over another four years. In other words, at any point, what someone held in “promised but not received” always exceeded “already received.” For ordinary employee RSUs it was harsher: 25% of the total vested after holding one year and six months post-IPO, with the remainder vesting monthly at 1/48. The earliest grants took four and a half years from IPO to fully vest. The earlier the grant, the longer the lockup. This wasn’t generosity—it was precisely calculated punitive retention.

One risk most people miss: golden handcuffs have a cliff when they expire. When that four-and-a-half-year vesting period ends, the employee’s account is clean and the company holds no new rope. Facebook’s design didn’t eliminate this risk, only delayed and smoothed it. The real solution is ensuring the next set of handcuffs locks on before the previous ones expire. The standard for evaluating any vesting arrangement is singular: on the day the last tranche vests, does the employee still have unvested future grants?

Tools Are Signals: Options Bet on Growth, RSUs Pay Compensation

What tool you use matters more than how much you grant.

Options operate on leverage. Employees share every dollar of appreciation above the exercise price, essentially making everyone bet on company growth. They’re most efficient during growth phases. Google granted approximately 28.5% of total shares outstanding in the three years before going public, averaging 7.1% annually, with over 99% as options. Exercise prices before 2002 ranged from a few cents to two dollars; the final 2004 batch was set at 70% of the IPO price.

But options deteriorate with rising valuations. After Microsoft’s 2007 investment valued Facebook at $15 billion, high exercise prices compressed the upside for employees, causing incentive efficiency to plummet. A high-valuation company issuing options is essentially issuing lottery tickets unlikely to pay out.

Google’s trajectory is the complete sample of this curve. Starting in 2007, they used a two-to-one option-to-RSU mix, balancing leverage and stability. After the 2008 financial crisis deepened in 2009, they shifted to nearly one-to-two, tilting toward stability. After 2014, they stopped issuing options entirely, using only RSUs. Post-IPO annual grants stabilized at 1.5% to 2% of total shares, with incentive costs around 5% to 6% of annual revenue. A tool’s retirement marks a company’s shift from betting on growth to paying wages.

One detail deserves separate mention. Facebook’s October 2007 mass switch to RSUs was primarily defensive: employees exercising early to avoid taxes and selling shares on secondary markets threatened to breach the 500-shareholder SEC threshold before going public. The RSU switch was forced by regulation. Retrospectively, this forced move perfectly matched their high-valuation phase needs. My assessment: correct tool selection is often forced by constraints. The difference is some companies learn to continuously reassess after being forced once; others remain stuck after being forced ten times.

A fair point for employees: many understand RSUs as “more stable,” but this intuition doesn’t hold for employees. Under US tax code, RSUs trigger tax liability upon vesting, potentially requiring taxes on depreciated stock. RSUs stabilize company-side accounting and dilution management. The first-principles question for tool selection is singular: do you want employees betting on company growth or receiving stock-denominated wages? The former requires leverage, the latter certainty. Mixing them creates neither.

When Markets Crash and Options Go Underwater, Character Shows

Between 2008 and 2009, Google’s stock fell 45%. Many post-IPO options had exercise prices above market price, becoming “underwater” options. For employees, incentive value dropped to zero; for the company, previously recognized incentive costs were wasted. This is the reality all option-issuing companies face in bear markets.

Google’s 2009 approach is worth copying, but the boundaries they set are more worth copying. One-time repricing: employees could exchange high-exercise-price old options one-for-one for new options priced near current stock price. Three simultaneous conditions: no quantity increase, old options canceled for new grants; new options reset waiting periods with extended vesting; the two founders, CEO, and non-executive directors were excluded from the program.

Each condition serves one purpose. No quantity increase protects shareholders from secondary dilution, since investors also suffered losses. Extended vesting prevents repricing from becoming an arbitrage window. Excluding management preserves the action’s fairness. Most companies do option repricing during bear markets, typically with the company unilaterally bearing extra cost while employees benefit at no cost—then everyone waits for rescue during the next price fluctuation. Google’s approach is an exchange: the company helps you stop losses, you pay with longer lockup.

Boundaries require clarity. This design has two implicit premises: long-term upward company value and accessible public market exits. Both premises face discounts in the Chinese market. IPO timing is uncontrollable, private company old shares lack liquidity, and what employees “see” may not convert to what they “get” for years. Real-world responses include cash-based long-term incentives and virtual dividends or discounted buybacks for long-term employee holders. My view: buybacks play the pressure-release valve role in this time machine. Without a release valve, “seeing always exceeds receiving” gradually transforms from incentive to arrears. Companies issuing equity should decide from day one where buyback funds will come from.

One number worth noting: pre-IPO, over 85% of companies repurchase vested shares from departing employees at prices reflecting gains but significantly below levels for current employees. Exit pricing design determines the retention structure’s integrity at the final moment.

The Pool Burns Faster Than You Think

Finally, resources themselves. Equity pools initially range from 10% to 20% of total shares, with actual pre-IPO grants typically landing between 10% and 15%. Usage pace averages 2% to 3% annually pre-IPO, with 60% to 70% of the total granted in the three years before IPO. The bulk of incentive resources gets spent in earlier stages.

These numbers point to an easily misunderstood fact. Total planning is a static problem; pace is dynamic. Insufficient pool size is rare; more common is burning too fast early: using half the pool at Series A, then having to compress percentages or dilute all shareholders again when recruiting executives at Series B. Later grants cost less—each valuation round requires fewer shares to retain the same person. But later grants risk losing people. You can issue early or issue aggressively, usually only one, with the other yielding. This tradeoff has no standard answer, only one hard constraint: when the pool runs dry, are the people the company most needs to retain still within their vesting periods?

Back to you as reader. If you’re receiving an offer, don’t focus solely on grant numbers. Ask “when does the last tranche of this grant vest, and what happens if I leave midway?” The vesting schedule matters far more than the total. If you’re designing incentives, treat equity compensation as temporal structure, not total compensation to divide. The test is one sentence: good design makes leaving at any point feel like a loss; bad design makes people feel square with the company only on vesting day.

This article’s methodological boundaries: the cases cover only Google and Facebook, both cycle survivors who became giants. Their parameters can’t be directly transplanted. What’s transferable is structure—the mismatch between granting and vesting, matching tools to stages, crisis remediation with boundaries, plus a pressure valve. Parameters must be recalculated using your own lifecycle.