Genius Sports: The Picks-and-Shovels Play Finally Has a Cash Register
The first full Legend quarter turns prediction-market optionality into $54 million of free cash flow—and management wants investors to start valuing it that way.
GENI · Earnings Call · 2026-11-05
Genius Sports: The Picks-and-Shovels Play Finally Has a Cash Register
Genius Sports reported a strong Q3 2026: revenue of $270 million beat its $260 million guide, adjusted EBITDA of $87 million beat the $85 million guide, and the adjusted EBITDA margin expanded to 32%, up 12 percentage points year over year. The company also generated $54 million of free cash flow, a 61% conversion from adjusted EBITDA, and raised full-year revenue guidance to $1.015–$1.035 billion and adjusted EBITDA guidance to $287–$297 million. The headline is a beat-and-raise. The more important change is the scoreboard: management is now asking investors to judge the company on free cash flow, not just revenue growth or adjusted EBITDA.
“From here on out, you will hear me reference free cash flow as an important metric to demonstrate the earnings power of the combined business.” — Mark Locke, Chief Executive Officer (CEO) · 2026-11-05 That is a meaningful shift for a company that spent years assembling rights, technology, and the Legend acquisition. The story is no longer 'can Genius assemble the assets?' It is 'can the assets convert into cash?' The first full quarter with Legend suggests the answer is yes—or at least that the machine is now pointed in that direction. The keyword earnings power has moved from investor-deck language to the P&L.
Prediction markets: the optionality gets a P&L
Prediction markets have been discussed on Genius calls since at least 2025, but Q3 is when the theme graduates from Q&A speculation to a revenue line with named customers. CEO Mark Locke called it “the biggest opportunity in our industry right now, and in financial markets more broadly.” — Mark Locke, Chief Executive Officer (CEO) · 2026-11-05 He pointed to Kalshi and Polymarket carrying a combined valuation of roughly $60 billion based on recent funding rounds. That is a striking number for a category that barely existed in the company's revenue base a few years ago.
Genius monetizes the prediction-market ecosystem at three points: customer acquisition through its media and Legend properties, trading through data sold to market makers, and settlement through official league data. Management says it has added more than 50 new customers this year across those stages. The pitch is deliberately platform-agnostic: if prediction markets take share from traditional sportsbooks, Genius still gets paid; if regulation pushes the activity back into sportsbooks, Genius already serves them. Locke put it bluntly:
There is no version of this that does not pay Genius.
The regulatory question is the obvious risk. Analysts pressed on whether a Supreme Court decision or state-by-state oversight could reverse the prediction-market model. Locke's answer was that the regulatory wrapper will evolve, but the underlying betting volume is durable: states are already taxing exchange fees or proposing transaction levies, and tribes in California have signaled interest in gaming regulation. The company's prediction markets exposure is real, but it is still not in the numbers: management explicitly said NFL prediction-market upside is “not in our guidance for this year, next year, or 2028.” — Mark Locke, Chief Executive Officer (CEO) · 2026-11-05 That is both the bull case and the discipline. Investors are being asked to pay for optionality that management refuses to underwrite. The same refrain appeared on the May call—“the upside isn't priced -- sorry, isn't in any of the guide” — Mark Locke, CEO · 2026-05-07—so the theme is not new. What is new is the revenue contribution and the customer count.
Legend and the Moment Engine: execution, not revelation
Legend is performing well. Management says integration is ahead of schedule on commercial, technology, and culture, and revenue synergies are arriving earlier than planned. The Kalshi and Polymarket deals are the visible proof points. Prediction.com, the new destination for prediction markets, is another: users who engage with its real-time trade feed are five times more likely to transact, according to the company. Media revenue rose 126% year over year, helped by Legend and by prediction-market and sportsbook ad spend.
The accounting deserves attention. CFO Bryan Castellani noted that more media revenue is now flowing through the self-serve platform on a net basis, which lowers reported revenue but raises margin. That is not a problem, but it makes organic growth harder to read. The Moment Engine added 279 advertisers in Q3 after 174 in Q2, bringing the cumulative base to roughly 450, with net revenue retention of 224% versus Q2. That is impressive early traction. It is also not a new theme—the Moment Engine has been in the company's keyword trajectory for several quarters, and prior calls mentioned 70 new customers and integration across 90% of the SSP/DSP market. The question is whether 224% NRR is an early-adopter effect or a durable land-and-expand curve.
The quiet pivot: from buying rights to harvesting them
One of the most important lines on the call was about rights costs. Castellani said data-rights costs are declining significantly as a percentage of revenue now that the new rights added in fall 2025 have lapped, and that long-term agreements give visibility into the next several years. Locke went further: the rights market has become rational, GeniusIQ is now embedded in rights negotiations, and “we don't actually need any more rights deals.” — Mark Locke, Chief Executive Officer (CEO) · 2026-11-05 That is a strategic pivot. For years Genius grew by acquiring rights—NFL, Premier League, Serie A, European leagues. Now the portfolio is largely locked to the end of the decade, and the margin story is cost leverage. The keywords rights fees and rights deals have been top-of-mind for investors through the acquisition phase; the current call marks a shift from land-grab to harvest.
The Q3 margin evidence is real: gross margin expanded to 34% from 25%, stock-based compensation fell to 6% of revenue from 16%, and the company produced $10 million of GAAP operating profit. If rights costs keep declining as a share of revenue, incremental betting and media revenue should flow through at high margins. That is the bridge to the 2028 targets of 35% adjusted EBITDA margin and at least 60% unlevered free cash flow conversion.
Balance sheet, buybacks, and the founder collar
Free cash flow was $54 million in Q3, and management expects more than $100 million across the second half, ending the year with roughly $260 million of cash. Net leverage is expected to fall to about two times by the end of 2026. Capital allocation priority remains debt repayment, not buybacks. When an analyst asked why not pay down debt faster, Castellani said the focus is delevering now, with optionality later. That is conservative and sensible given the leverage profile.
Locke also disclosed a personal transaction: he sold about 800,000 shares, roughly 4% of his vested holding, and arranged a collar over another 4.4 million shares at his own cost. He said the cash is for family and tax planning and that he remains substantially invested. The collar is a governance nuance rather than a thesis changer, but founder confidence is part of the equity story, so the disclosure matters.
What matters from here
The Q4 guide implies an adjusted EBITDA margin of about 34.5%, which would be the highest Q4 margin in the company's history. But the raise came mostly from the Q3 beat, not from a Q4 uplift—management said Q4 assumptions were essentially unchanged. Brazil's closure mutes some upside but is not material to the 2028 guide. The bigger swing factor is regulation: if prediction markets are forced into a state-by-state sportsbook model, volumes may not disappear, but the economics could shift. Genius says it wins either way; the market will need more proof than one quarter.
The quarter's real message is that Genius is no longer just a rights aggregator with a prediction-market call option. It is a cash-generating infrastructure layer with a new KPI and a 2028 target. The first full Legend quarter says the machine works. The next four quarters will say whether it compounds. For a small-cap name—market cap around $1.6 billion—that is a legitimate re-rating case, provided free cash flow keeps converting and the regulatory ground does not shift under it.