Brian Nichols is the co-founder of Angel Squad, a community where you’ll learn how to angel invest and get a chance to invest as little as $1k into Hustle Fund's top performing early-stage startups

Key takeaways

• Databricks raised a $5.24 billion Series L in December 2025 at a $134 billion post-money valuation and $190 per share, led by Insight Partners, Fidelity, and J.P. Morgan Asset Management.
• Secondary units trade near $219.74 as of early June 2026, roughly 16% above the last round price and up about 97% over twelve months.
• The company crossed a $5.4 billion revenue run rate in February 2026, growing 65% year over year, with more than $1.4 billion of that coming from AI products.
• That puts the secondary at roughly 29x run-rate revenue, with demand outstripping available supply by 3.3 to 1.
• Databricks does not allow direct stock transfers, so your only access is indirect: special purpose vehicles or forward purchase contracts. Both carry risks most buyers underprice.

Databricks crossed a $5.4 billion annual revenue run rate in February 2026. A year earlier it was around $3 billion. For a company that did roughly $1 million in revenue in 2015, that is one of the steepest curves in enterprise software, and the secondary market has noticed. Units that changed hands near $111 last summer now trade around $220.

So the appetite is real. The question is whether the price still makes sense at this altitude, and what you are actually buying when you click "invest" on a Databricks secondary. Let's get into the numbers, then the structure, because the structure is where most people lose money they did not need to lose.

What the Numbers Actually Say

Start with entry price versus the last round. The December 2025 Series L priced shares at $190. The secondary, per Notice's algorithmic consensus, sits near $219.74. That is about a 16% premium to the most recent primary price. Worth pausing on: you are paying more than the institutions paid six months ago, for stock that is harder to sell and comes with less information.

The per-share march tells the story plainly. Series J priced at $92.50 in December 2024. Series K hit $150 in September 2025. Series L landed at $190 that December. The secondary tacked on another 16% on top. That is roughly 2.4x on the primary in about a year, plus a secondary premium. For more on why a higher entry price quietly eats into your outcome, our breakdown of venture capital return multiples is the piece to read before you commit.

Trailing multiple. Notice pegs Databricks' real-time market cap at about $155.67 billion. Against a $5.4 billion run rate, that is roughly 29x. Against the $134 billion last-round valuation, the institutions bought in at about 25x. So the secondary is asking you to pay up from where the smartest money in the room just transacted.

Forward multiple. If revenue keeps compounding at even 55% to 60%, run rate lands somewhere north of $8 billion over the next year. On that figure, the same market cap works out to high-teens forward, call it 18x to 19x. That is the bull case in one line: today's nosebleed multiple becomes a merely expensive multiple if growth holds.

Peer comparison. Hold that 29x next to Notion, another AI-accelerated private name, which trades closer to 17x ARR on the secondary. Databricks commands the premium because it is growing faster and because demand for its shares is outrunning supply. But here is the sober frame: public data and analytics names generally trade well below these private secondary multiples. You are paying a private-market markup for illiquid stock, and that markup only pays off if the growth story stays intact through an eventual listing. For context on how the public window actually opens and closes, see our note on the IPO market for angel investors.

Revenue trajectory. The trend line is the cleanest part of the thesis. Roughly $3 billion in early 2025, $4.8 billion by the third quarter, $5.4 billion by February 2026. Revenue per employee climbed from about $303,000 in 2024 to roughly $398,000 on a trailing basis across more than 15,000 people. The AI segment alone is a $1.4 billion run-rate business now, which matters because it is the part of Databricks competing in the fastest-moving corner of the market.

Profitability and margins. Here is the honest gap. Databricks does not publish its margins, and the last public net income figure anyone saw was a $44 million loss back in 2020. At a $5.4 billion run rate that number is ancient history, but the point stands: you are buying a company that has not shown you its bottom line. Treat the margin picture as unknown, and price that uncertainty in.

The Demand Story Is Doing a Lot of Work

Notice reports investor demand outstripping the supply of available shares by 3.3 to 1. That imbalance is most of why the secondary trades at a premium to the primary. When more buyers want in than there are sellers, price drifts up regardless of what a spreadsheet says the company is "worth."

This is exactly how our very own Hustle Fund GP, Elizabeth Yin, frames pricing. As she puts it, valuations are about supply and demand: the supply of a given round or tranche, and the demand from investors chasing it. A 3.3-to-1 imbalance is that dynamic running hot. It is a genuine signal that smart people want exposure. It is also the thing that can reverse fastest. The day supply loosens, through a tender, a secondary program, or simply a few large holders deciding to take chips off the table, that premium compresses. You want to underwrite the business, not the queue of people standing in line behind you.

The Hype Question

Databricks sits at the center of the AI trade, and AI is about as hyped as a sector gets. Elizabeth's long-standing test for hyped markets is whether a company is genuinely "10x different and 10x better" than the alternatives, not just riding the wave. Databricks has a real claim here. Its Lakehouse architecture and its land-and-expand motion inside large enterprises are sticky in a way a thin AI wrapper is not. The $1.4 billion AI run rate is being sold into accounts that already run their data on the platform.

That is the steelman. The bear case is that the company competes with the deepest pockets in technology, and that "AI products" growth invites a customer acquisition fight where everyone is spending to win the same logos. Both things can be true. The reason to keep following the late-stage playbook carefully is that at this size, your return depends less on the product winning and more on the entry price you paid for a near-certain winner.

How You Would Actually Buy It (And Why That Matters)

Databricks does not allow direct stock transfers. Read that twice, because it changes everything about the trade. You cannot simply buy a share and hold it in your name. Your exposure has to come through a vehicle, and the two common ones are special purpose vehicles and forward purchase contracts. If you are newer to how these structures pool capital into a single late-stage position, our SPV strategy guide covers the mechanics. Two risks deserve specific attention.

Risk one: forward purchase contracts are synthetic exposure. A forward is not a share. It is a contract in which a current shareholder promises to deliver shares (or their economic value) to you at a future date, often at an exit. You are not on the cap table. You are holding a promise. If the seller defaults, or the company blocks the eventual transfer, or the seller's own arrangement falls apart, you are an unsecured creditor of a counterparty rather than an owner of Databricks. That counterparty risk is the whole ballgame, and it is the part that gets waved away in a hot market. Know exactly who is on the other side of your forward and what happens if they cannot deliver.

Risk two: second and third-layer SPVs stack fees and bury the truth. In demand-heavy names like this one, the vehicle offered to you is frequently not the one holding the stock. It is an SPV that bought into another SPV that holds a forward that references the shares. Every layer adds a management fee and a slice of carry, and every layer makes it harder to see your real cost basis. By the time a "$220 Databricks unit" reaches you through two or three nested vehicles, you may be paying a meaningfully higher effective price than the headline, and you may not be able to see it. Before you wire, ask how many layers sit between your check and the actual shares, and what each layer charges. If nobody can answer cleanly, that is your answer. The same scrutiny you would apply to deal terms and protective provisions on an early-stage round applies double here.

What Has to Go Right, and What Has to Go Wrong

For the bull case: growth stays above 50%, the AI segment keeps compounding, and Databricks lists or gets acquired at a market cap that makes 29x look cheap in hindsight. Plausible. The business is excellent.

For the bear case: growth decelerates, the multiple compresses toward public comps, your entry premium evaporates, and the indirect structure clips your upside with fees while exposing you to counterparty risk on the way down. Also plausible, and underpriced by a market that is currently bidding 3.3 to 1.

The asymmetry is not as friendly as it looks, because you are entering at a premium to the last round in a structure that takes a cut. None of that makes it a bad investment. It makes it an investment you should size and structure deliberately rather than chase.

The Bottom Line

Databricks is one of the genuinely great private software companies, and the secondary demand reflects that. The catch is that "great company" and "great entry" are different questions, and at roughly 29x run-rate revenue through a layered, indirect structure, the second question deserves more of your attention than it usually gets.

This is the kind of deal where being inside a serious community pays for itself, because the structuring details (how many SPV layers, who the forward counterparty is, what the real fee load looks like) are exactly what gets missed when you go it alone. That is a big part of why Angel Squad exists. It is a community of more than 2,500 investors across 50-plus countries who have collectively put over $30 million into 70-plus startups, with a strict no-a-holes policy and access to the top 1% of deal flow. Hustle Fund is an early-stage fund, but Squad members see the full spectrum, from pre-seed all the way through pre-IPO names like this one, with the context to actually vet the structure before they commit. You can learn more at hustlefund.vc/squad.

Frequently Asked Questions

What is the current Databricks valuation? Databricks was last priced at $134 billion post-money in its December 2025 Series L. On the secondary market, Notice's real-time consensus puts its market cap closer to $155.67 billion as of early June 2026, reflecting demand above the last round price.

What is the Databricks secondary share price? Secondary units trade near $219.74 as of early June 2026, roughly 16% above the $190 Series L price. That figure is an algorithmic consensus built from secondary transactions and reference data, not a guaranteed execution price.

How much revenue does Databricks generate? Databricks reported a $5.4 billion annual revenue run rate in February 2026, up about 65% year over year, with more than $1.4 billion coming from its AI products.

Can I buy Databricks stock directly? No. Databricks does not permit direct stock transfers. Accredited investors access it indirectly through special purpose vehicles or forward purchase contracts, each of which carries fees and counterparty considerations.

Is Databricks profitable? Databricks does not publicly disclose current margins or net income. The most recent public figure was a net loss in 2020. At today's scale that is dated, but profitability should be treated as undisclosed.

When will Databricks IPO? There is no confirmed timeline. The company has repeatedly signaled it is in no rush to go public, so any secondary purchase should be underwritten as a multi-year hold with real liquidity risk.