DraftKings sells contracts where it cannot sell bets
DraftKings' map has no overlaps. The company offers its online sportsbook in 27 states, plus Washington D.C. and Puerto Rico, and sells contracts on sporting events in another 18, including California, Texas, Florida, and Georgia. The remaining five only receive free contests. No state appears on two lists at once, and this architecture supports three distinct arguments that the company needs to maintain simultaneously.
The commercial slogan is true. The "Take Your Game Anywhere" campaign, which launched on August 31, relies on the brand reaching all 50 states, something Chris Mendez, creative director at Translation agency, attributed to the combination of the sportsbook and predictions. Stephen Miraglia, the company's senior communications director, explained it clearly to Gambling Insider: where there's a sportsbook, there are no predictions, and where the sportsbook is not allowed, contracts and daily fantasy appear.
The first problem this design solves is sports-related. The NFL confirmed this week that advertising for prediction markets is prohibited in its games, stadiums, club agreements, and player endorsements, and that its betting partners can only advertise approved products, without generic advertising. DraftKings, one of the three partners signed on August 27 along with FanDuel and Fanatics, maintains that it keeps the 50-state campaign outside the league's windows. Since the prediction product has no customers in states where the sportsbook is advertised, the restriction costs it almost nothing.
The second front is regulatory. On August 25, the company launched a prediction campaign specifically targeting California, Texas, Georgia, and Florida, the four largest states without legal online sports betting, offering USD 200 in bonuses after an initial USD 5 transaction. Jason Robins, its CEO, was explicit that he chose them for that reason. The message isn't to bet here instead of there, but rather that here, there's nothing else.
The third is financial, and the thesis here is that the two products don't cannibalize each other. In the first-quarter report, Robins wrote that they see no perceptible impact of predictions on the betting business, and that internal and third-party data showed a minimal effect on industry handle, concentrated in low-margin bets. The second-quarter report, from August 6, went further: it spoke of massive new customer acquisition in states without regulated sportsbooks, about a 1 percent customer overlap between its sportsbook and the largest prediction market operator in states where it does operate, and its own estimate that between 80 and 90 percent of the consumption volume of these markets in said states comes from professional syndicates and institutional traders. The overlap data comes from card deposit information processed by Carbon Arc; the 80 to 90 percent figure is an internal calculation.
The problem with these numbers is that they describe the map as it was drawn. If the prediction product is confined to states where the company does not have a sportsbook, a low overlap in sportsbook states is a consequence of the design and not proof of what the customer would do if they had both options available.
Volume, meanwhile, is soaring. Robins reported that annualized volume increased from USD 2.3 billion in April to USD 11 billion in July, company figures annualized from a single month, implying about USD 900 million handled in July. Corporate slides break down this total into USD 3.6 billion annualized consumer volume and USD 7.4 billion from market makers, meaning actual retail operations are around USD 300 million monthly. Integrating predictions into the main application cut the customer acquisition cost for that product by more than 80 percent in April.
The margin tells another story. Since the second quarter, the company has reported a combined line item that groups sportsbook and prediction revenues against a volume that combines handle and operations. Volume increased 14.5 percent year-over-year to USD 13.14 billion, but revenue fell 10.6 percent to USD 892 million, and net margin dropped from 8.7 to 6.8 percent. The company attributes this to customer-favorable sports results, particularly the Knicks' championship, which hit hardest in its largest state, and the World Cup group stage, in addition to promotional reinvestment in acquisition. In the earnings call, Robins quantified the sports results as about USD 80 million in headwinds and stated that, normalized for both factors, revenue grew 10 percent. What he didn't say is how much predictions contribute to that blended figure. Sportsbook handle alone grew 11 percent, so about four points of volume growth correspond to prediction market operations, at a revenue rate that the company does not disclose and described only as lower per customer than that of its sportsbook.
There's a line from the first quarter that should not be overlooked. The company reported that preliminary third-party data suggests prediction customers lose their money faster than sportsbook customers. It presented this as an argument for trust and consumer protections. It is also a description of the product it is advertising in the four largest states where no regulated sportsbook exists.
The entire scheme rests on a legal gamble. On August 10, the company self-certified nine categories of American football contracts on DKeX, its proprietary market announced on June 26, covering game winner, handicaps, totals, and player statistics: the full menu of a sportsbook, for sale precisely where the NFL partner cannot sell the actual product. The league, which told the Commodity Futures Trading Commission that prediction market rules fall far short, has in return the guarantee that none of this will appear in its broadcasts.
That is an arrangement between a league and its largest betting partner, not an answer to the fundamental question. The 18 states on the prediction list are, by definition, those where no legislature has authorized sports betting. Whether a federally regulated contract on a Cowboys game handicap constitutes a sports bet in Texas is what the Ninth and Third Circuits resolved differently, and what New Jersey asked the Supreme Court to define. For now, DraftKings' map assumes the answer is no.
Tags: DraftKings, prediction markets, DKeX, US betting regulation