Understanding stock options.
The Airtable acquisition story.

The mountain: how a $11.7B company sells for $1.3B

Airtable did almost everything right. Real product, real revenue — about $480 million a year, still growing 20% when it sold. What collapsed wasn't the business. It was the price: in the 2021 zero-interest-rate frenzy, investors paid roughly 100× revenue. When rates rose and AI tools reset the category, no public market would ever clear that mark again.

Notice the shape: 54% of every dollar Airtable ever raised came in at the very top — $735M at $11.7B in December 2021. That single round is the villain of every chapter that follows, because those investors didn't just buy shares. They bought shares with rights that sit in front of yours.

Meet Maya, employee #212

Maya is a hypothetical (but very typical) senior engineer who joins in September 2020, right after the Series D valued the company at $2.585 billion. Her offer letter includes the line every startup employee remembers forever:

Fifteen months later, the Series F prices the company at $11.7B — about $23.40 per share. Maya opens a spreadsheet and multiplies: 40,000 × $23.40 = $936,000. She starts browsing real estate listings.

That number is not wrong, exactly. It's just the answer to a question nobody asked: "what would Maya's shares be worth if they were the same shares the VCs bought?" They aren't. Which brings us to the single most important chart in this guide.

Four prices for the same share

At any moment, a startup share has several simultaneous "prices." They can differ by 10× or more, and each one governs a different decision in your life:

Maya's $936,000 used the blue number. Her taxes would be computed with the yellow one. Her real chance to sell was the orange one. And what she'd ultimately be paid was the green one — nine times smaller than the number in her spreadsheet.

The tax trap: paying real money on paper gains

In January 2022 — peak euphoria — Maya has 20,000 vested options. Conventional wisdom says exercise early to start the long-term capital-gains clock. So she does. Here is what actually happens to her bank account:

1. Exercising is a purchase

She must buy the shares: 20,000 × $1.60 strike = $32,000 in cash, wired to the company, for stock she cannot sell.

2. The IRS taxes the spread — even though no money arrived

The gap between the 409A value and her strike — ($8.50 − $1.60) × 20,000 = $138,000 — counts as income the moment she exercises. Because these are ISOs, it's Alternative Minimum Tax (AMT) income: roughly $38,000 of tax. Had they been NSOs, it would be ordinary income taxed immediately at ~35–45%. Either way: a five-figure tax bill on a gain that exists only in an appraisal.

3. The 90-day guillotine

If she ever leaves the company, she typically has 90 days to exercise vested options or lose them — meaning the choice between "walk away from your equity" and "write a large check for illiquid stock" often arrives exactly when you're changing jobs.

Maya is now out $70,000 of real cash — a year of savings — holding paper that yields nothing, can't be sold, and whose "value" is an appraisal. This is the position thousands of employees at 2021-vintage unicorns found themselves in. The ones at Airtable were comparatively lucky. At Fast, Convoy, or InVision, the same trade ended at zero.

Secondaries: the only real price before the end

A company's last-round valuation is a negotiated artifact — it can stay frozen at $11.7B for years simply because nobody raises a new round. But shares kept trading anyway. In late 2021, some employees sold in tender offers near the peak. By early 2026, Airtable stock changed hands on secondary markets at an implied $4 billion — a 66% discount to the sticker price — months before any deal was announced.

The secondary market knew. It usually does. When real buyers with real money price your company at a third of its headline valuation, that's not noise — that is the market's honest estimate, and it predicted the $2.25B exit far better than the cap table did.

Two rules fall out of this:

Rule 1 — a tender offer is information, not just liquidity. If the company organizes a chance to sell at a good price, the people organizing it believe the price is good for buyers of certainty. Selling 20–50% is rarely a mistake; at Airtable, an employee selling in the 2021 tender at ~$20 locked in roughly 7.7× the eventual exit price.

Rule 2 — never buy secondaries above the preference-adjusted price. Whoever bought at the $4B secondary mark in early 2026 lost ~45% in months, because they paid a common-share price for a company whose top $1.2B of exit value belonged to someone else.

Exit day: the waterfall

August 2026. Bending Spoons pays $1.285B for the business; adding Airtable's ~$1B of unspent cash, about $2.25B flows to shareholders. But it does not flow evenly — it flows in order. Investors with liquidation preferences choose whichever is larger: their money back, or their ownership share. When the exit is below the valuation they paid, they take the money back — first.

Read that left bar carefully: the investors who put in $1.19B at the top of the market got exactly $1.19B back — a 0% return, but a protected 0%. Employees have no such protection. Common stock absorbs the entire fall from $11.7B to $2.25B before preferred loses a cent. That asymmetry is the whole reason "the company is worth $11B" never meant your shares were.

This is why the single best diligence question to ask any startup is: "How much has been raised, at what valuations, with what preferences?" Airtable's answer — $1.36B raised, mostly at 5–10× the eventual exit — told you the ending in advance.

Three endings for Maya

Same company, same talent, same luck — the only variable is what each version of Maya did with her options. The differences are brutal:

The uncomfortable moral: outcome tracked selling discipline, not loyalty or effort. Maya B didn't outsmart anyone — she just treated a tender offer at a 100×-revenue valuation as what it was: a chance to convert a forecast into money. Maya A round-tripped six years of paper wealth to roughly break-even. Colleague C paid $170,000 for the privilege of losing $118,000 — with no bad decisions except trusting the sticker price.

What to actually do

1. Value your grant at the exit-common price, not the headline.

   A decent haircut: mentally value options at 25–40% of the preferred-price math, less if the company raised a lot recently. If the number still excites you, great — you'll be surprised upward.
2. Know your three numbers before you sign or leave.

   Strike price, current 409A, and total capital raised with preferences. Companies will share the first two always and the third if they respect you. A refusal is also information.
3. Model the tax bill before exercising — it's a purchase plus a tax event.

   ISOs: the 409A-to-strike spread hits AMT. NSOs: it's ordinary income immediately. Exercising early (when spread ≈ 0, with an 83(b) election) avoids the tax but converts it into purchase risk — see Colleague C. Never exercise more than you can afford to lose entirely.