Tender Offers and Employee Liquidity at Agent Unicorns: How GaaS Staff Actually Cash Out

Table of Contents

• Why Liquidity Is the Quiet Story at Agent Unicorns
• Tender Offer Mechanics, Plainly
• Why GaaS Companies Run Tenders Differently
• The Valuation Trap: Primary Price vs. Tender Price
• Taxes, 409A, and the Mistakes That Cost Real Money
• How to Read a Tender Offer Document
• What Founders Get Wrong About Liquidity Programs
• Insights Most People Overlook
• References

Why Liquidity Is the Quiet Story at Agent Unicorns

Talk to anyone who joined a fast-scaling agent company in the last two years and you'll hear the same thing: the offer letter promised life-changing equity, and the cap table says it's true, but the bank account hasn't noticed. That gap, between paper wealth and spendable wealth, is the defining tension of working at a GaaS unicorn right now.

The reason is structural. These companies are raising at valuations that would have been late-stage a decade ago, but they're doing it at Series B or C, sometimes with eighteen months of revenue history. They are nowhere near an IPO. A company worth $4 billion on paper that won't go public until 2029 leaves a four-year window where employees hold illiquid stock, owe taxes on vested RSUs in some structures, and watch the market re-rate agent companies up and down with every model-pricing announcement.

Tender offers are how that window gets survivable. They're not a sideshow. For most agent-unicorn employees, a well-run tender is the single most important financial event between joining and a hypothetical exit, and the failure to run one is a leading cause of regretted early departures. This piece sits inside our broader GaaS funding cluster; if you want the upstream context on how these companies get over-capitalized in the first place, the mega-round phenomenon in agent infrastructure sets the stage for why liquidity pressure builds the way it does.

Tender Offer Mechanics, Plainly

A tender offer is a structured, company-sanctioned buyback of shares from existing holders. In the typical late-stage version, an incoming investor (or a syndicate) agrees to purchase a fixed dollar amount of existing stock alongside, or shortly after, a primary financing round. The company sets the rules: who's eligible, how much each person can sell, and at what price.

A few mechanics worth internalizing:

• It's almost always capped. You rarely get to sell everything. Companies set a percentage cap, commonly 10% to 25% of an individual's vested holdings, to keep employees invested and to keep the demand pool within the dollars the buyer committed.
• Eligibility has tenure gates. Most programs require shares to have vested and the employee to have been there a minimum period, often one to two years. Recent joiners frequently get nothing in the first tender.
• The price is negotiated, not market-discovered. There's no order book. The buyer and the company agree on a per-share price, which may be at, below, or occasionally above the last primary round price.
• It's a discrete event, not a standing window. Tenders open for a few weeks, then close. Miss the election window and you wait for the next one, which might be a year out, or never.

The structured-secondary format has become standard enough that platforms like Carta's overview of company-run tender offers treat it as routine late-stage plumbing. The key distinction from a wildcat secondary sale, where an employee finds a buyer on their own, is governance: a company tender is sanctioned, priced uniformly, and usually the only way to sell without tripping transfer restrictions in your stock agreement.

Why GaaS Companies Run Tenders Differently

Agent companies share the broad late-stage playbook, but three features of the GaaS business model bend how their liquidity programs work.

Compressed Timelines Force Early Tenders

A traditional SaaS company might run its first tender at Series D, five or six years in. Agent unicorns are running them at Series B, sometimes within thirty months of founding, because they've hit unicorn valuations that fast. That compression is a direct consequence of how aggressively the market values these companies, a dynamic we cover in why agent startups command premium valuations. The upshot: employees face liquidity decisions with far less operating history to judge the company's durability.

Margin Volatility Makes Pricing Contentious

Here's the GaaS-specific wrinkle. An agent company's gross margins ride on model-inference costs it doesn't control. When a foundation lab cuts token prices, margins expand overnight; when usage spikes on a per-outcome contract that underpriced compute, margins compress. That volatility shows up in tender pricing. Buyers underwriting a secondary at an agent company demand a discount that a comparable SaaS secondary wouldn't carry, because the forward margin picture is genuinely murkier. The revenue-quality question around whether usage revenue is durable is the exact uncertainty that gets priced into a tender discount.

Talent Wars Make Liquidity a Retention Weapon

Comp at top agent startups is brutal, and the people most likely to get poached are exactly the ones holding the most valuable early equity. Founders increasingly run tenders not because investors pushed for them but because a competitor's recruiter is dangling a liquid offer. A tender becomes a retention tool: let your senior research and infra staff take some chips off the table so they don't leave for the company that already did. McKinsey's work on how generative-AI talent is reshaping compensation and retention underlines how acute that pressure has become at the frontier.

The Valuation Trap: Primary Price vs. Tender Price

The single most misunderstood thing about tenders is the relationship between the primary round price and the price you actually get.

When a company announces a $3 billion round, that headline reflects the primary preferred shares the new investor bought, shares loaded with liquidation preferences, anti-dilution protection, and other rights that common stock (what employees hold) doesn't have. The tender price for common stock is routinely set at a discount to that preferred price, sometimes 20% to 40% lower, because common is genuinely worth less.

So the math that feels like a bait-and-switch is often just the preferred-vs-common spread doing its job. But not always. Watch for two darker patterns:

1. A tender priced well below the last primary round can signal an internal markdown the company hasn't announced. If new money came in at $80/share but the tender is offered at $45 and the preferred-common spread doesn't explain the gap, the company's own read on its value may have softened. This is adjacent to the down-round risk that haunts over-funded agent startups.
2. A tender used to reset the 409A valuation can cut both ways. A secondary at a real arm's-length price can become a data point that lowers, or raises, the company's 409A, which in turn moves the strike price on future option grants.

Treat the tender price as the only price that's real to you. The primary headline is marketing for everyone except the holders of preferred stock.

Taxes, 409A, and the Mistakes That Cost Real Money

This is where smart engineers lose six figures to avoidable errors.

Long-term vs. short-term capital gains. If you exercised options early and held the shares more than a year before the tender, your gain may qualify for long-term capital-gains treatment. If you're selling shares you only recently acquired (or selling via a same-day exercise-and-sell), you can get hit with short-term rates plus ordinary income, a difference that can exceed 20 percentage points on the federal side alone.

QSBS is the giant lever nobody mentions. Qualified Small Business Stock can exempt a large portion of gains from federal tax if the shares were held more than five years and the company qualified as a small business when issued. Agent unicorns scale so fast that the five-year clock and the gross-asset thresholds get tricky, but employees who exercised very early, when the company was tiny, may be sitting on a substantial QSBS exemption they'll forfeit if they tender too soon. The IRS guidance on the Section 1202 small-business-stock exclusion is the authoritative starting point, and this is a "talk to a tax advisor before you click sell" situation if there ever was one.

RSU structures change everything. Many later-stage agent companies switched from options to double-trigger RSUs. Those don't get taxed until both vesting and a liquidity event occur, and a tender can, depending on structure, constitute that second trigger, creating a tax bill on shares you didn't even sell. Read your specific plan documents; the answer is not generic.

The cluster's piece on the secondary market for GaaS startup shares covers the off-platform alternatives, but the tax analysis there is the same minefield.

How to Read a Tender Offer Document

When the packet lands, most people skim the price and the cap and sign. Read these five things first:

• The eligibility and cap table. Confirm your exact eligible share count and per-person cap. Companies make errors here; vesting-cliff edge cases get miscalculated.
• The price and how it was set. Look for whether an independent valuation backed the price or whether it was simply negotiated with the buyer. The former is more defensible.
• The repurchase-rights and ROFR language. Some tenders quietly bundle in waivers or expansions of the company's right of first refusal on your remaining shares. You may be signing away future flexibility.
• Tax-withholding mechanics. Especially with RSUs, understand exactly what gets withheld and whether you'll owe more at filing.
• The information rights you're getting, or not. A well-run tender comes with enough financial disclosure to make an informed decision. If you're asked to sell into an information vacuum, that itself is a signal.

The asymmetry is real: the buyer has the company's full financials and you have a four-page summary. Close that gap as much as the documents allow before you elect.

What Founders Get Wrong About Liquidity Programs

A note for the people on the other side of the table, because the cluster's audience includes operators.

Founders consistently under-run liquidity. They treat a tender as a favor to be doled out sparingly, when it's actually one of the highest-leverage retention and culture levers they have. The failure modes are predictable: waiting too long (until a competitor forces it), capping too aggressively (so the program feels insulting), or running it so opaquely that employees assume the worst about why the price is what it is.

The companies that handle this well do three things. They run tenders on a predictable cadence so employees can plan. They communicate the preferred-common spread honestly so the discount doesn't read as a betrayal. And they size the buyer commitment so the cap is meaningful, letting someone sell 3% of their stake is theater. As a16z has argued in its writing on employee equity and the modern liquidity landscape, staying-private-longer only works if the liquidity infrastructure keeps pace, and agent companies are staying private during the most compressed value-creation curve software has ever seen.

Insights Most People Overlook

1. The first tender is a tell about the second. How a company runs its inaugural liquidity program, the cap, the price honesty, the disclosure, predicts how it'll treat employees at exit. A founder who runs a generous, transparent first tender is signaling something. One who runs a stingy, opaque one is also signaling something. Weigh that signal into your stay-or-go decision more than the dollar amount.

2. Tender demand is a cleaner durability signal than the primary round. Primary investors buy a story and a preference stack. Secondary buyers in a tender are buying naked common stock with no downside protection, on a forward margin profile they had to underwrite cold. If sophisticated secondaries are eager to buy an agent company's common at a thin discount, that's a stronger vote of confidence in revenue durability than any lead-round headline. Sparse tender demand at a steep discount is the inverse warning, and it often precedes the funding-slowdown indicators the rest of the market notices months later.

3. Selling your full cap can be the right move even when you believe in the company. The standard advice to "hold your winners" was written for public markets where you can rebuy. In a private agent unicorn with margin volatility and a multi-year exit horizon, concentration risk is enormous and you cannot diversify a position you can't sell. Taking your full allowed cap off the table, even while staying employed and bullish, is frequently the rational play, not a betrayal of conviction.

4. The 409A reset is the part founders forget to weigh. Running a tender at a real arm's-length price can drag the 409A valuation in directions that affect every future hire's option strike. A tender that prices common low to clear demand can inadvertently lower strike prices for incoming talent, a hidden recruiting upside, or, in a hot market, ratchet 409A up and make future grants pricier to employees. Founders rarely model this second-order effect.

5. "Liquidity" without disclosure is a transfer of risk, not a gift. The most under-discussed tender pattern is the company that offers a sell window but withholds the financials needed to value the shares. Employees selling blind are doing the buyer a favor: they're offloading shares the buyer can price with full information and the seller can't. A real liquidity program comes with real disclosure. If it doesn't, the generosity is illusory.

References

• Tender Offers Explained, Carta
• QuantumBlack / Generative AI Insights, McKinsey
• About Publication 550, Investment Income and Expenses (Section 1202 QSBS), IRS
• Andreessen Horowitz, a16z