Startup tender offer or wait for the IPO?

What happened at ElevenLabs, and why should you care?

A tender offer is a company-run buyback window that lets employees sell part of their vested equity at a fixed price without waiting for an IPO — and ElevenLabs just ran one of the largest ever. The AI voice company said this week it had closed a $300 million employee tender that valued the business at $22 billion, co-led by Wellington Management and T. Rowe Price. That's double the $11 billion it reached in February, when Sequoia led a $500 million Series D. It was also the company's second liquidity event in about a year: a $100 million tender priced it at $6.6 billion in September 2025. New backers including EQT, Goldman Sachs, GIC and Sapphire Ventures bought in for the first time.

If you work at an AI startup — or you're about to accept an offer from one — this isn't just industry trivia. Tender offers have gone from rare favour to standard practice. The tender market grew to roughly $35 billion in 2025, close to the scale of the entire US IPO market that year, and the average gap between one tender and the next at a given company shrank from around 899 days in 2022 to about 132 days by 2025. US venture secondary volume passed $100 billion. Companies that once told employees to wait for the bell now run a liquidity window every year or two, because that's cheaper than going public.

So you face two separate decisions, and people tend to mash them together. The first is whether to sell now, how much, and what the tax bill looks like. The second is how to use a published tender price the next time a recruiter says "the equity is the real upside" — because a recent tender gives you a per-share number that isn't a founder's hope or a stale 409A. Get the first decision right and you diversify; get the second right and you can add five or six figures of grant value to an offer. The rest of this article works through both.

What does a startup tender offer actually give you?

A tender offer gives you partial liquidity at one fixed price, on terms the company sets in advance — it is not access to a market. The company, or an outside investor syndicate as in ElevenLabs' case, states the price per share, the total dollars available, who may sell, how much each person may sell, and the deadline. You decide how many eligible shares to tender. That's the extent of your control: no haggling on price, no negotiating your own cap, no promise the window reopens next year. Treat it as an option that expires, because that's precisely what it is. The board can also amend or withdraw the offer before it closes, so nothing is real until the cash clears your account.

Eligibility is where most people get a nasty surprise. A two-year tenure cliff is the most common rule, and some issuers set it at four years, so recent joiners are simply left out. Only vested shares can be tendered — that's universal — and some offers cover only shares you've already exercised and own outright, not unexercised options. Roughly six in ten tenders include former employees, usually within two years of leaving. Founders, board members and early investors are often carved out to keep the paperwork manageable. Caps on how much of your vested holdings you can sell commonly land between 15% and 25%, sometimes expressed as a dollar ceiling instead.

The mechanics are more formal than they look. You'll receive an Offer to Purchase that can run 40 to 80 pages, covering price, eligibility, payment, recent financials, risk factors and tax disclosure, with an election window that typically runs about 20 business days once non-accredited holders are involved. Read the financials section — for most employees it's the only real look at company performance they ever get. If employees collectively ask to sell more than the buyers want, allocations are cut back pro rata, so plan for the possibility that you sell less than your cap. Pricing often tracks the current 409A value, though a tender run alongside a raise can price higher, as ElevenLabs' did.

Tender offer vs IPO: what does waiting actually cost you?

Waiting costs you time and concentration risk, and the time bill has grown brutal. Recent US tech listings have gone public roughly 12 to 13 years after founding, up from eight to ten years two decades ago, and median time to exit for late-stage companies now sits around 15 years. Nearly half of today's US unicorns raised their first round in 2016 or earlier and still haven't listed. If you joined a four-year-old company, "wait for the IPO" may quietly mean waiting until the late 2030s — through a bear market, a strategy pivot, and several rounds you don't control. Meanwhile your single largest asset is unsellable and undiversified.

The structural risks compound. Every new round dilutes you and stacks more preferred stock above your common shares, so a flat or down exit can pay investors first and leave employees with a fraction of the headline number. Most venture-backed companies never IPO at all; they get acquired quietly, wind down, or grind on. And secondary liquidity is concentrated in a handful of perceived winners — one OpenAI tender last October accounted for about 6.2% of full-year US secondary volume. Outside that small group, shares aren't merely discounted, they're often untradeable. A live tender at your employer is a privilege, not a baseline you can count on repeating.

Here's the honest counterweight: sellers into ElevenLabs' $6.6 billion tender last September watched the price hit $22 billion twelve months later. That's roughly 3.3x foregone on whatever they sold, and no amount of diversification theory makes that feel good. So the real question isn't "sell or hold" — it's what share of your net worth you can afford to leave riding on one private company's next three years. My view, unfashionable in startup Slack channels: take the full cap in your first tender. You keep 75–85% of your upside, you stop making life decisions around an illiquid asset, and you learn what your paperwork actually says while the stakes are modest.

Which tax and timing questions should you ask before you sign?

Three clocks decide what a tender actually nets you, and you need to check all three before the election deadline. For incentive stock options, a sale qualifies for favourable treatment only if it happens more than two years after grant and more than one year after exercise; miss either and you have a disqualifying disposition, where the spread between exercise price and fair market value at exercise is taxed as ordinary income. The second clock is the plain one-year long-term capital gains holding period. The third is the five-year clock for qualified small business stock, and it's the one that costs people the most money when they ignore it.

QSBS rules changed with the 2025 tax law, so the answer depends on when your shares were issued. Stock issued before 4 July 2025 follows the old rules: a full five-year hold and a $10 million per-issuer cap on excluded gain. Stock issued after that date gets tiered treatment — 50% exclusion at three years, 75% at four, 100% at five — with the exclusion rising to $15 million or ten times your adjusted basis, whichever is greater, and the company gross-asset threshold raised from $50 million to $75 million. Note the asymmetry: buyers who acquire your shares in a tender generally get no QSBS at all, because the stock must come directly from the company. If you must sell early, a Section 1045 rollover into new QSBS can defer the gain.

Then do the boring arithmetic. If you exercise options specifically to sell into the tender, the sale is a disqualifying disposition by definition, so model the ordinary-income hit and the withholding before you elect, not after. Non-qualified options are taxed on the spread at exercise, and the fair market value at exercise — not your strike — becomes your cost basis. Ask the equity team three questions in writing: what's withheld and at what rate, when the cash settles, and what tax forms you'll receive. My rule: a tax clock should override diversification only when the anniversary is weeks, not years, away. Otherwise you're paying real risk for a maybe-deduction.

How do you use a live tender price to negotiate a new offer?

A recent tender price is the single most credible per-share number a candidate can put on the table, so use it as your valuation basis and make the company do the maths with you. When a recruiter quotes equity, ask: "At the price from your last employee tender, what is this grant worth, on a fully diluted basis, per year of vesting?" That one sentence converts a vague share count into an annual dollar figure both sides can argue about honestly. If the company has never run a tender, that's information too — it means your shares have no demonstrated price and no path to cash until an exit, which justifies asking for more of them or more cash up front.

Before you counter, collect six facts: the latest 409A value, the preferred or tender price from the most recent transaction, total fully diluted shares outstanding, what percentage your grant represents, your strike price, and the post-termination exercise window. Then apply the haircut practitioners use — roughly 20% for common versus preferred, plus about 20% for illiquidity — so you're comparing apples to apples against a public-company offer. Know what's actually movable: the strike price isn't, because tax rules pin it to the 409A on your grant date, but share count, the cliff, the vesting schedule, the exercise window and sign-on cash all are. Ask when the next 409A refresh lands, too.

Leverage comes from alternatives, not arguments. Two written offers move more equity than the cleverest spreadsheet, which means the fastest way to improve your equity package is to get further along in more processes. That's a CV problem before it's a negotiation problem: if the file doesn't clear automated screening, you never reach the conversation where a tender price matters. Run your CV through a free AI CV analysis to see your score, ATS compatibility and category breakdown before you start applying — and if you're an engineer weighing AI-startup offers, check that your CV passes the first screen with your stack written where parsers actually read it.

Frequently asked questions