Three years of “profitability over promos” just ended, and the ad slate this NFL season is the evidence. DraftKings and FanDuel are each pouring $200-300 million-plus into prediction markets ahead of kickoff, a combined bet projected to cost them more than $500 million in adjusted EBITDA, while BetMGM has simultaneously launched the most unapologetically brand-first campaign the category has produced in years. Those two things are happening in the same three-week window, from operators who all report to shareholders demanding the same thing: disciplined, provable ROI on marketing spend. They can’t both be right about what this season calls for, and that’s the actual story.
The short answer: Kalshi and Polymarket have reopened an acquisition channel that didn’t exist in the “discipline era” playbook, prediction markets in states where sports betting isn’t legal, and DraftKings and FanDuel have decided the land-grab is worth re-running. That single decision has split the category. Operators exposed to the prediction-market fight are spending like it’s 2021 again and calling it customer acquisition; operators that aren’t are free to run pure brand campaigns, like BetMGM’s, that would have looked reckless to greenlight if they were the ones facing a nine-figure EBITDA hit this quarter.
The macro backdrop: why NFL season 2026 broke the truce
Total NFL wagering this season is projected at roughly $29.5 billion, flat versus last year. Flat handle at this scale, with a new and well-capitalised category of competitor entering the market, means growth has to come from somewhere, and right now it’s coming out of sportsbooks’ acquisition budgets rather than organic market expansion.
The NFL has picked a side. It renewed its sportsbook partnerships with DraftKings and FanDuel and added Fanatics as a third official partner for 2026, while explicitly declining any deal with prediction market operators. That’s not incidental, it hands state attorneys general a data point in their own disputes with Kalshi and Polymarket over whether federally-regulated “event contracts” are functionally sports betting operating outside state licensing. DraftKings and FanDuel have reportedly gone further, leaving the American Gaming Association over this exact issue, under what’s been described as shareholder pressure not to cede market dominance in states where sports betting still isn’t legal.
Here is the commercial logic behind the $200-300 million-plus each operator is spending: DraftKings says customer overlap with prediction market users in legal sportsbook states is only around 1%, and that 80-90% of prediction market volume comes from professional traders rather than recreational bettors. The pitch to the board isn’t “we’re cannibalising our own base”, it’s “we’re buying a legal front door into states we can’t otherwise enter.” CEO Jason Robins has staked the company’s public thesis on the claim that predictions-customer lifetime value will end up comparable to sportsbook-customer LTV. That’s the number this NFL season is actually testing, underneath all the ad creative.
Not every operator thinks the trade is sound. PENN Entertainment CEO Jay Snowden used the word “irrational” to describe where competitive spending could go this season, warning of an “arms race” mentality creeping back into a category that had spent three years disciplining itself out of exactly that behaviour. Whether that’s genuine concern about margin compression across the product, or PENN signalling that its own restraint is a choice rather than a lack of capital, it puts every marketing team in the category in the same bind, match the spend and defend acquisition ground in a fight that isn’t really theirs, or hold discipline and watch two much larger competitors buy a multi-decade customer relationship on the cheap.
What the ad campaigns reveal about the marketing-commercial split
Put this season’s biggest NFL campaigns side by side and they stop looking like five variations on the same celebrity-endorsement brief. They’re five different bets about where growth comes from next, and where each operator sits on the prediction-market fight explains a lot about which bet they’re making.
| Operator | Campaign | Face | What it’s actually selling |
|---|---|---|---|
| DraftKings | “Take Your Game Anywhere” | Kevin Hart, Nick Jonas | Cross-sell across sportsbook, casino, predictions, fantasy, lottery and horse racing via a single super-app, retention economics, not just reach |
| Fanatics | 360-degree campaign / Jersey Drop | Creator partners, Jay-Z’s 40/40 Club | Fan-experience positioning and a $5m prize pool to buy fast credibility as the NFL’s newest official partner |
| BetMGM | “Make It Legendary” | Jon Hamm | Pure brand repositioning, agency Highdive says on the record it’s “not about direct conversion” |
| Caesars | Loyalty-tied campaign | Eli & Peyton Manning, Garth Brooks | Celebrity banter format tied directly to the loyalty programme, the legacy playbook, still running |
| Hard Rock Bet | Social/commercial campaign | Post Malone | Cultural authenticity, positioned against the “exclusivity” perception in gambling advertising |
BetMGM’s is the cleanest brand play in the set, and it’s the one that makes the most sense once you notice BetMGM has no direct exposure to the prediction-market spending war. CMO Casey Hurbis called it “a reimagining of what BetMGM stands for”; Highdive’s Mark Gross went further, framing the whole platform as “storytelling, connection and creating a brand that truly speaks to today’s players, not at them.” That’s a marketing team buying brand equity with no near-term conversion target attached, a far easier internal pitch when your board isn’t simultaneously approving a nine-figure loss elsewhere in the P&L.
DraftKings and Fanatics are running the opposite logic underneath similar-looking celebrity creative. “Take Your Game Anywhere” exists to move customers across DraftKings’ full product stack inside one login, a retention and cross-sell mechanic dressed as a brand campaign, and the more defensible spend when the same marketing organisation also has to justify $200-300 million going into prediction markets this quarter. Fanatics, as the newest of the NFL’s three official partners, isn’t optimising for retention at all, its job is rapid, credible awareness, which is why the $5 million Jersey Drop prize pool and the Jay-Z 40/40 Club activation lean on real cash mechanics rather than pure storytelling. Caesars and Hard Rock Bet sit between the two poles: Caesars ties its Manning-brothers-and-Garth-Brooks format directly to loyalty-programme conversion, while Hard Rock’s read on Post Malone as “authentic representation” rather than a traditional spokesperson is a tacit admission that star wattage alone has a ceiling.
That ceiling shows up in the data. A 2021 YouGov study found only 12% of Americans consider celebrities effective endorsers for gambling products specifically, and that the demographic most swayed by celebrity gambling ads (55+) is roughly twice as susceptible as the 24-40 cohort every one of these campaigns is actually built to reach. That’s the format the entire category defaults to for NFL season, measurably weakest with the audience it’s targeting, which makes the split between BetMGM’s brand bet and DraftKings/Fanatics’ commercial mechanics less about creative taste and more about which operators can afford to run a campaign that isn’t required to move a KPI this quarter.
Does the math actually work?
Every one of these campaigns eventually answers to the same spreadsheet. Standard category economics put customer acquisition cost somewhere between $250 and $750 (conservative CAC in the current market), against a two-to-three-year lifetime value of $1,200 to $1,800, a return of roughly $4-7 for every $1 spent in an established market. That math is what let operators spend $1.2-1.3 billion each on sales and marketing in 2024 (26.5% of revenue for DraftKings, 22% for FanDuel’s US segment) and still tell shareholders the category had matured into discipline. It’s also what let category-wide TV ad units fall 17% year-on-year through 2024 even as total spend (about $666 million, still under 1% of national TV ad spend) held flat, operators buying the same reach more efficiently, shifting mix toward performance and affiliate channels that deliver verified depositors instead of impressions. Caesars had already shown what real discipline looks like at the extreme, cutting marketing spend by more than 75% in an earlier cycle once it had banked share.
That’s the consensus prediction markets just broke. The $4-7x benchmark assumes a customer who can eventually be monetised through a full-margin sportsbook product, parlays, in-play markets, loyalty-driven repeat handle. Prediction market customers, by DraftKings’ own account, are 80-90% professional traders operating in a product structured more like a financial exchange than a bookmaker’s book. That’s a different margin profile entirely, which is exactly why Robins’ claim, that predictions-customer LTV will match sportsbook-customer LTV, is the number worth watching this quarter, not the size of the marketing spend itself. If he’s right, the $200-300 million each operator is spending is a customer acquisition cost with a familiar payback curve, just booked through a different product wrapper. If he’s wrong, it’s loss-leading market-share defence dressed in acquisition-spend language, and a much harder story to tell on the next earnings call.
Read the ad campaigns again with that lens and the pattern holds: DraftKings’ cross-platform push is built to maximise how many product lines a single acquired customer touches, which is the only sound strategy if you’re not fully certain the predictions LTV thesis holds on its own. BetMGM’s Jon Hamm campaign can afford to skip that question entirely, because it isn’t the one carrying the bet.
What this means for your own acquisition budget
None of this is really about celebrity casting. It’s two departments, brand and commercial, making incompatible bets about the same NFL season, and getting away with it because they sit inside operators with very different exposure to a fight most of the category isn’t even part of. If you’re building or defending acquisition spend this quarter and you’re not DraftKings, FanDuel or a direct prediction-market competitor, the mistake is benchmarking against their spend level. The more useful exercise is re-running your own CPA and LTV assumptions against a category where the two largest players have just accepted a lower near-term ROI for long-run market access, because that decision resets the competitive cost of acquisition for everyone else, whether or not you ever touch a prediction market yourself.
Is your team’s NFL-season budget actually being justified against this new baseline, or is it still being defended with last year’s numbers?
Leave a Reply