Allan Stone, Intelitics, on prediction markets: ‘Most of the gap is marketing

In an interview with European Gaming, Allan Stone, chief executive officer at Intelitics, explains what Kalshi’s World Cup run exposed about operator marketing, which parts of an operator’s measurement model stop working when a prediction market trader arrives, and why he thinks most of the gap between the two categories is self-inflicted.

Key findings

• On what Kalshi got right: ‘Fifteen people beat companies with a hundred times their headcount.’
• On acquisition costs: ‘If your 2026 budget was built assuming last year’s competitive set, it’s already wrong.’
• On measurement: ‘Right now, most operators are trying to fit a new animal into an old cage.’
• On lifetime value: ‘That’s one data point pretending to be a trend line.’
• On the regulatory argument: ‘Fix the regulatory gap, and there’s still a marketing problem sitting underneath it.’

What Kalshi’s World Cup run actually proved

European Gaming:You posted recently that sportsbook marketing teams should study Kalshi’s World Cup run. What did Kalshi get right that operators did not?

Allan Stone: Fifteen people beat companies with a hundred times their headcount. Sit with that for a second, because it’s the whole story.

Their trading volume during the tournament came in higher than most sportsbooks’ entire handle, with almost six million app downloads. They landed on a brand momentum list next to companies people have known their whole lives, ahead of the network that actually aired the games. Beating the broadcaster on brand recall during the broadcaster’s own tournament isn’t luck.

The audience is the part that operators should sit with the longest. 

• Kalshi’s female user base grew at almost double the rate of its male users through the tournament, and by late June, women made up close to a third of their traders.
• Sportsbooks have been stuck in the low 20s on that number for years. A lot of these were people who’d never placed a bet on anything before in their life.

The usage pattern backs it up. Sportsbook activity spiked early in the tournament and cooled off fast, the usual pattern. Kalshi kept building all the way through the knockout rounds.

I’m not calling any of this permanent yet. Both Kalshi and Polymarket saw sports activity drop hard the moment the World Cup tournament ended, which tells me some of that was tourist traffic chasing a moment, not a new customer base showing up for good. 

But the question I’d put to every operator this week is simple: 

‘If Kalshi turns even a third of those new traders into people who show up for elections, weather, or whatever’s next, does the rest of the industry have an answer for that? I don’t think it does yet.’

The new bidder in every auction

EG:Prediction markets are now bidding against sportsbooks for the same users on the same channels. What has that done to acquisition costs this year?

Allan Stone: Ask BetMGM. Their own CEO said the quiet part out loud this earnings season, that prediction markets are pulling marketing dollars out of the same channels sportsbooks live in, pushing acquisition costs higher and stretching out payback on new cohorts. That’s not me reading tea leaves. That’s a public company telling investors why guidance moved.

DraftKings and FanDuel are now projecting more than half a billion dollars combined in lost EBITDA this year from building and marketing their own prediction products. PENN’s CEO called what’s coming before football season an arms race, and used the word irrational to describe it.

‘If you’re not one of the three companies in that fight directly, your CPMs go up anyway. You’re bidding in the same auctions as multiple public companies plus a well-funded prediction market that just raised at a valuation higher than DraftKings, and none of them are optimizing for your payback window.’

Cheap capital already stopped covering for bad unit economics this year. Now there’s another bidder in every auction who doesn’t need your math to work in order to keep spending. 

If your 2026 budget was built assuming last year’s competitive set, it’s already wrong.

Where the measurement model breaks

EG:A prediction market trader behaves differently from a sportsbook depositor. Which parts of an operator’s measurement model stop working when that user shows up?

Allan Stone: Most operators built their measurement around a bettor’s rhythm. Deposit, place a few wagers around Sunday’s slate, maybe a parlay, wait for the next slate. Your LTV cohorts, your predicted-value windows, your CRM triggers, all of it assumes that shape.

A prediction market trader doesn’t move like that. 

• They check a probability the way you’d check a stock price, all day, across a news cycle instead of a game clock. 
• Their money might sit as working capital across a dozen small positions instead of one stake on one outcome. 
• The signals that usually tell you someone’s a real bettor and not a bonus hunter, early session frequency, bet size, sport selected, don’t map cleanly onto someone trading probability on a Fed decision or an award show.

‘Attribution breaks first. Your CRM was built to say this person is a sports bettor or a casino player. It has no clean bucket for someone who found you through a market on the World Cup and checks in daily without ever placing what your system recognises as a wager.’

If your data model can’t describe the behaviour, it can’t value it. Right now, most operators are trying to fit a new animal into an old cage.

Why nobody has a twelve-month number yet

EG: Does anyone yet know what a prediction market user is worth over 12 months, or is the industry still guessing?

Allan Stone: Nobody has it. Anyone telling you different is guessing with confidence, which is worse than just guessing.

I flagged this right after the World Cup. Both Kalshi and Polymarket saw sports activity fall off hard the second the tournament ended. Chasing a moment isn’t the same as building a customer base, and right now, most of what gets reported as growth in this category is still moment-chasing.

Jason Robins said publicly that DraftKings believes it can eventually get predictions customers to the same lifetime value as sportsbook customers. Notice the word believes. That’s a projection, not a measured cohort. Kalshi rolling out a VIP tier called Platinum earlier this year is the same story from the other side. You don’t build a white-glove retention program for users you’ve already sized up. You build it because you’re trying to find out who your whales actually are.

A real twelve-month LTV number needs a cohort that lived through a full calendar, deposits, withdrawals, a slow month, a reactivation push, and at least one tentpole event without letting it carry the whole number. Prediction markets at this scale have had exactly one World Cup. That’s one data point pretending to be a trend line.

‘The honest move for 2027 budgeting is to treat any LTV figure you hear from this category as a hypothesis, not a benchmark, until someone’s watched a cohort for a full year without a tentpole event propping it up.’

Affiliates are repricing the shelf space

EG: Affiliates have moved quickly into prediction market traffic. How are they pricing it, and what does that do to sportsbook CPAs?

Allan Stone: Affiliates go where the money moves fastest, and right now that’s prediction markets. Kalshi is largely running this on promo codes, the person who owns the code gets paid when someone new funds an account and actually trades. 

Polymarket’s version runs on wallet activity, on-chain, and rewards flow automatically. Neither one looks like a mature sportsbook rate card yet. It’s CPA, revshare, and hybrid deals getting stitched together in real time, because nobody running these programs has years of cohort data to price it properly.

‘That’s the exact mistake I’ve been warning operators about for years, just showing up in a new category. You can’t price a partner deal off volume when you don’t know what the volume is worth yet.’

What it does to sportsbook CPAs is simple. 

• Affiliates have finite inventory, their best placements, their trusted lists, their highest-intent audiences. 
• If a prediction market will pay for that placement today with less friction and a newer, more exciting story for their audience, sportsbooks have to pay more to keep the shelf space they used to get by default. 
• You’re not losing affiliates. You’re bidding against a category that doesn’t need the deal to make sense yet.

What transfers to Europe, and what stays home

EG:Nine European regulators have coordinated against unlicensed prediction market platforms, and the UKGC treats them as betting intermediaries. If Europe licenses the product rather than blocks it, does the US marketing playbook transfer?

Allan Stone: Partially, and the part that doesn’t transfer is the important part.

Nine regulators across Europe, Belgium, France, Germany, Italy, the Netherlands, Poland, Portugal, Spain, and Switzerland, have already coordinated against unlicensed prediction platforms. The UK Gambling Commission went further and said a platform like Kalshi or Polymarket would fall under the legal definition of a betting intermediary if it operated in Great Britain, the same category as a betting exchange like Betfair, which has been regulated there since roughly 2000.

If that becomes the framework, prediction markets in Europe stop being a grey-zone growth story and become a licensed product competing on the same terms as everyone else. That changes what winning looks like.

The US playbook right now works partly because the ambiguity itself is the growth engine. Move fast, stay culturally relevant, let the regulatory question stay open while you scale. 

A licensed European market removes that lever entirely. You’re not out-hustling anyone into your app anymore. You’re fighting the same onboarding, payments, and KYC battle regulated operators have always fought, the same one I’ve written about with offshore books beating regulated ones on user experience, not on being unregulated.

The instinct to move fast and operate like a newsroom, that part travels anywhere. The advantage of nobody being quite sure what you are yet, that one stays home.

Regulatory gap or marketing gap

EG: Operators say prediction markets have an unfair regulatory advantage. How much of the gap is regulatory, and how much is that they are simply being out-marketed?

Allan Stone: Both are real, and operators need to stop hiding behind the first one to avoid admitting the second.

The regulatory gap is legitimate. Prediction markets sit under federal CFTC oversight while sportsbooks fight state by state for every license. Sixteen states already took action against sports-related event contracts last year, and the AGA puts the lost state and tribal revenue at a billion dollars. That’s a real structural asymmetry, not a talking point.

But if I’m being straight, most of the gap is marketing, not regulation.

‘Fifteen people out-executed some of the best-funded, most experienced brands in this industry over a six-week tournament, out-downloaded them, out-recalled the network that aired the games, and pulled in a female audience sportsbooks have been trying and failing to reach for years. Nothing about the CFTC made that happen. Better marketing did.’

If I had to put a number on it, I’d say closer to a third regulatory and two-thirds self-inflicted. Operators had years of runway, brand recognition, and bigger budgets, and still got beaten on cultural relevance by a company a fraction of their size. Fix the regulatory gap, and there’s still a marketing problem sitting underneath it.

One change before the 2027 budget

EG: For a marketing team setting its 2027 budget, what is the one thing you would tell them to change now?

Allan Stone: Stop building the budget like prediction markets are a footnote and sportsbooks are the whole competitive set. They’re not anymore, and that assumption is the most expensive one a marketing team can carry into 2027.

• Build one LTV-based budget instead of a media plan with a threat assessment stapled to the back. 
• Get real player value signals into the ad platforms before your next big calendar moment, not during it. 
• And keep a reserve you can move in days, not a quarter, because the operators who lost ground this year weren’t out-budgeted. They were out-reacted.

‘If 2026 taught this industry anything, it’s that the team willing to move like a newsroom beats the team that planned like it was still 2019.’

About Intelitics

Intelitics is a marketing intelligence platform for betting and gaming companies, covering marketing attribution, predictive lifetime value, performance marketing and partner management. The company is based in Costa Mesa, California. Allan Stone is its chief executive officer.

The post Allan Stone, Intelitics, on prediction markets: ‘Most of the gap is marketing, not regulation’ appeared first on European Gaming Industry News.