Image credit: Source: company disclosures and earnings materials.
A ready-made audience was supposed to be the cheat code for customer acquisition in US betting. ESPN Bet spent about $1 billion proving it is not. PENN Entertainment terminated the ESPN partnership in November 2025 and rebranded to theScore Bet, ending the highest-profile bet on audience-led growth the industry has run.
The theory was clean. Acquisition is the single largest cost in sports betting acquisition, so start with tens of millions of fans and the cost falls away. The graveyard of 2025 and 2026 says attention and customers are not the same thing.
The most expensive audience in sports
PENN did not lack reach. It rented the biggest one in American sport.
The 2023 deal had PENN paying ESPN $1.5 billion over ten years, roughly $150 million a year in cash, plus warrants for about 31.8 million PENN shares worth around $500 million, per the terms disclosed at signing. ESPN Bet launched that November across 16 states with ESPN's audience, app integration and on-air talent behind it. The internal target was around 20% market share.
It peaked near 7% in December 2023, slid to about 3.2% by May 2025, and sat near 3% when PENN pulled the plug, per Sportico. Over roughly two years, PENN booked about $1 billion in cumulative adjusted losses on the venture before rebranding to theScore Bet in December 2025 and cutting senior roles.
The most-watched sports brand in the country could not turn viewers into a top-three book. That is the data point every growth team should sit with.
Reach is not the retreat it looks like
ESPN Bet is not an isolated failure. The whole audience-first cohort has pulled back.
- FanDuel TV, the market leader's own media channel, is being wound down by the end of 2027, with a 60% workforce cut starting in June 2026. Even the number-one operator is retreating from owning a broadcast audience.
- Betr, the Jake Paul and Joey Levy media-first micro-betting startup, raised at a $375 million valuation in 2024, then pivoted. Its real-money sportsbook now runs in roughly two states while it leans into free-to-play, launching a social casino across 34 states in March 2026.
The pattern is consistent. A media audience is cheap to reach and expensive to convert. Eyeballs on a broadcast do not open a funded wagering account, pass identity and geolocation checks, and bet through a losing month. That last part matters most: retention, not acquisition, drives the economics, and a fan who came for free content churns the moment the boosts stop.
The one model that works owns wallets, not eyeballs
Fanatics is the exception, and the exception proves the rule.
Fanatics converts a commerce relationship, not passive attention. Its roughly 95 million to 100-million-strong customer base comes from selling merchandise and collectibles, people who have already handed over a card and a shipping address. Its rewards currency, FanCash, is earned on bets and redeemable for jerseys and memorabilia, and CEO Matt King has said the company can send a signed jersey to a bettor at "relatively low cost for us, but a huge" draw for the fan. King's framing is deliberate: "We're a sports brand, not a gaming brand," building "Spotify in a market dominated by iTunes," he told iGaming Business.
The results are real but modest. Fanatics reached roughly 5% share by the end of 2024 and around 6.8% of handle across the 20 states one operator analysis tracked, helped by buying PointsBet's US business for about $225 million. Its betting unit turned over roughly $300 million against $8.1 billion in group revenue in 2024, with commerce and collectibles profits funding the push.
The difference from ESPN is the nature of the relationship. Fanatics started with transactions, not impressions. A customer who already buys from you is a warmer lead than a viewer who watches you, and the loyalty currency gives a reason to stay that a broadcast never did. That is the same logic behind loyalty tiers replacing the welcome bonus: the durable asset is a paying relationship, not a marketing reach figure.
What actually lowers acquisition cost
Strip out the audience story and the winners look ordinary. FanDuel and DraftKings each hold north of 30% share, and they got there with product, scale and disciplined paid acquisition, not a captive broadcast audience.
Their media moves are support acts, not the strategy. DraftKings owns VSiN and struck a marketing deal with NBCUniversal in 2025; both are cheaper distribution, not a growth engine. The engine is a better app, faster markets, and a machine for turning a funded account into a habit. Where a captive audience does help, it helps at the margin and only when it comes with a transaction, which is why Stake's streamer-led distribution works where ESPN's broadcast did not: the creator drives a deposit, not just a view.
The commission math points the same way. As acquisition costs stayed brutal, affiliate deals shifted from flat CPA toward revenue share, tying payment to players who actually keep betting. Nobody who has priced US acquisition believes a cheap top-of-funnel audience is the answer anymore.
Attention is not a customer
The audience shortcut was the most seductive idea in US betting growth. Buy or build a big enough following, the logic went, and you skip the acquisition war. ESPN, the ultimate test case, ran it with the deepest audience in the market and roughly $1 billion to spend, and finished with about 3% share and an exit.
The operators winning are the ones treating acquisition as what it is: a per-customer cost that only pays back over a retained, transacting relationship. Fanatics converts buyers. FanDuel and DraftKings convert product and scale. ESPN tried to convert viewers, and viewers, it turns out, mostly just watch.
Related: Loyalty Programs Are Replacing Free Bets in US Betting | Retention, Not Acquisition, Now Drives iGaming Growth | iGaming Acquisition Shifts From CPA to Revenue Share | Stake Built a Marketing Machine Out of Casino Streamers