Dreamstime/Yobro10
August was a good month for online casinos in some states, but not so good in others.
Michigan online casinos saw a collective increase of 5.3% in gross gaming revenue (GGR). The total was $97.2 million, or $87.6 million in adjusted revenue after deducting promotional expenses. That’s a new record for the state. Both New Jersey and Pennsylvania have had $100 million-plus months, and Michigan looks likely to get there before the year’s out.
New Jersey went the opposite direction, seeing its online casino revenue drop 4.2%, or 4.6% for overall iGaming with online poker included. (Michigan does not separate out online poker revenues, so no figures for casino only are available).
Pennsylvania also declined, but to a lesser degree. Its online casinos shed 1.9%, though strong gains for online poker due to the arrival of WSOP.com mitigated the damage. Total iGaming was down only 1.2%.
Delaware online casino revenue was up significantly in August, while West Virginia’s was down slightly. These small states contribute very little to the national total, however.
On aggregate, then, the national market held close to steady despite state-to-state variations. Put together, the five states generated about $314 million in GGR, compared to $326 million in July, a drop of 3.7%. Small fluctuations like this have been the name of the game for the past six months or so, after what was a rather dramatic year from 2020 through early 2021.
That’s not to say nothing has changed over the summer, however. Though the total has held steady, the battle for market share has been drifting in favor of the traditional casino companies and away from the sports-focused newcomers.
All figures cited in this article come from data supplied by state regulatory authorities, with most calculations performed by Online Poker Report.
FanDuel and DraftKings feel pressure from both sides
Market share is a tricky thing to measure in most states. Regulators report the numbers by license, but in New Jersey, Pennsylvania and West Virginia, each license is shared by multiple brands.
Michigan gives us a unique opportunity to study the market share battle due to its one-skin-per-license rule. All the major brands are there, and the top operators – BetMGM, DraftKings and FanDuel – are the same Big Three that dominate nationally. For convenience, we can consider the next three – BetRivers, Golden Nugget and WynnBet – to be the second tier, and the remaining nine smaller operators to make up the third tier.
Here’s how market share has evolved from February to present. It’s evident that both DraftKings Casino and FanDuel Casino have been losing share in a consistent fashion. Both BetMGM, the market leader, and the second tier sites have been the beneficiaries of that.
Since February, BetMGM has expanded its market share from 34% to 41%, while the second tier sites collectively grew from 11% to 18%. At the same time, the combination of FanDuel plus DraftKings has dropped from 39% of the market to less than 28%.
Third tier sites gained a fair bit in August, thanks in part to the rebranding of William Hill Casino to Caesars Casino. Revenue for the Grand Traverse Band of Ottawa and Chippewa Indians, which host the rebranded casino, rocketed from under $296,557 to $822,344 as a result of the change, showing just how much power there is in a brand.
Data courtesy of the Michigan Gaming Control Board
Why are FanDuel and DraftKings losing market share?
The downside to looking at Michigan numbers is that they don’t go back very far. The market only launched in late January this year, making February the first useful data point. We therefore can’t see whether this is a temporary phenomenon, the continuation of a larger trend, or the first half of an annual cycle.
There’s reason to suspect there’s at least annual component to it, however. DraftKings and FanDuel have a lot in common. Mostly importantly, they’re both sports-focused companies that have begun cross-selling their customers on the casino vertical.
Sports betting revenue is extremely, consistently seasonal for obvious reasons. The exact cycle varies from country to country, depending on which sports are most popular. In the US, however, it peaks in February and March, thanks to the double-whammy over the Super Bowl followed by March Madness. After that, it drops off sharply, only climbing again in the fall when the NFL season gets underway again.
Casino revenue generally isn’t quite as seasonal, but may be for those two. Less sports betting means fewer cross selling opportunities. That most other operators have been growing their revenue at the same time suggests that there is underlying organic market growth masked by the losses of the sports companies.
Testing this hypothesis should be easy. The NFL season has resumed this month. If the shrinking casino share of DraftKings and FanDuel is indeed related to sports seasons, then they should begin bouncing back sharply when September numbers become available. That should also mean faster growth for the market as a whole.
For more information on sports betting revenue, see coverage at our sister site Legal Sports Report.
New Jersey online gambling revenue – Aug. 2021
Although New Jersey licenses are the most cluttered with brands, it looks as if the Garden State shows a similar trend in market share.
Almost all online casino license holders in the state lost revenue in August. The exceptions were Borgata and Caesars. The former mostly consists of BetMGM and its own Borgata Casino brand, with Pala Casino and Party Casino playing second fiddle. The latter’s most important skins are its own Caesars Casino and WynnBet, plus Harrah’s and 888. It’s consistent with Michigan that those two should have gained.
Meanwhile, the heaviest losses were for Hard Rock, Golden Nugget and Resorts. Golden Nugget hosts FanDuel’s casino product and its affiliated Stardust brand, while Resorts is home to DraftKings.
New Jersey online poker tumbled a whopping 20.4%, far worse than the 4.2% lost by the online casinos. That’s all from Caesars-owned WSOP.com, as its online bracelet series came to an end. Competitors PokerStars and Partypoker US Network (including BetMGM Poker) gained a little as players who’d moved to WSOP for the series switched back to their usual sites.
Pennsylvania online gambling revenue – Aug. 2021
On the surface, the Pennsylvania online casino numbers look as if DraftKings and FanDuel might have bucked their downward trend in other states. Valley Forge, which is home to FanDuel, held even, while Penn National, which hosts DraftKings among other casinos, lost 4.8%, which is small compared to some other licenses.
However, both those licenses saw a big drop in table game revenue, offset by large gains for slots. The core audience for the sports companies has shown a preference for table games in the past. Thus, a closer look at the numbers suggests that we are seeing a drop in sports-to-casino crossover, offset by gains for the pure casino segment.
Indeed, the contribution of slots to total GGR in PA was 78.3% for August, the highest level since July 2020.
Pennsylvania also saw a lot of movement among smaller brands. Caesars, PlayLive! and TwinSpires all saw huge gains, while Parx, Mohegan Sun and Wind Creek all took a beating.
Finally, August was the first full month of operation for WSOP’s poker room. It has established itself with about 25% of the Pennsylvania online poker market, while PokerStars holds 60% and the BetMGM/Borgata tag team making up the remaining 15%.
West Virginia online gambling revenue – Aug. 2021
West Virginia online casino revenue was amazingly flat in August. The Mountain State reports revenue weekly, not monthly, but comparing the total for the past four weeks to the four which preceded, the difference is less than 0.3%. Attempting to estimate the monthly total by counting partial weeks gives a different result, showing a drop of 4.9% from July.
The Greenbrier had the best month of the three active licenses, up 7.5%. However, it’s impossible to tell whether that’s due to FanDuel or BetMGM, which both operate on that license. Based on other states, the natural assumption would be that BetMGM increased its revenue, while FanDuel slumped by a smaller amount, but there’s no way to know for sure.
BetRivers, operating on the Moutaineer license, picked up 4.7%, while DraftKings, on the Charles Town license, lost 10.5%, wiping out those other gains and keeping the market even.
Delaware online gambling revenue – Aug. 2021
Delaware is a small and non particularly competitive market. Though there are three casinos offering online games, all three are operated by the state lottery, using technology from 888. Since there’s no real battle for market share, the three tend to track each other fairly well.
All told, the market revenue increased 19.2%, shared by all three casinos. Even that substantial gain wasn’t enough to push GGR past $1 million, a feat the state has only accomplished once before. That was in May 2020, while retail casino gaming was shut down due to the pandemic. For August 2021, the total was $945,512, so it may get there again if September is another good month.
Monthly total iGaming revenue by state
Here’s how the various state markets have evolved over the past month and past year.
| State | August ’21 iGaming GGR | Change (m/m) | Change (y/y) |
|---|---|---|---|
| New Jersey | $113,230,701 | -4.6% | +29.0% |
| Pennsylvania | $103,245,569 | -1.2% | +54.8% |
| Michigan | $97,197,617 | +5.3% | N/A |
| West Virginia | $5,034,486 | -4.9% | +210% |
| Delaware | $945,512 | +19.2% | +16.9% |
Some important notes about these figures:
- Revenue given is gross gaming revenue, or our best estimate of such, regardless of what a given state reports. GGR is defined as wagers collected minus winnings paid out, and ignores any promotional expenditures.
- Monthly growth is calculated as a daily average to correct for varying number of days in each month.
- iGaming revenue includes both online casino and online poker, but not sports betting, instant online lottery or any retail gambling products.
- West Virginia reports revenue on a weekly basis, so monthly totals are estimated by splitting weeks based on number of days falling within the month. This may introduce random error on the order of a few percent.