I’d start with people who want to pay and can’t. In many of the markets where payers are growing fastest, people pay with Pix, UPI, or OXXO more than with cards. And in some markets, local rules make it hard to save a card, so buyers enter their details every time.

Next, I’d stop losing buyers who already tried to pay. When a good customer’s purchase is declined, Roblox loses the sale and sometimes the customer. Network tokens, routing through the orchestration layer I describe in the Platform section below, and smarter retries are usually the fastest wins in payments. The Plus subscription supercharges the revenue too since it’s a monthly subscription, so a renewal that fails on an expired card loses a subscriber who potentially never chose to leave.

On web, every extra screen in the funnel causes churn and reduces purchases. I’d make checkout one tap with a saved card or Apple Pay, and ask for extra authentication only when a purchase looks risky. I know Roblox changed the checkout experience to cater to kids who charge their parents’ credit cards with the time-bound loader purchase button that fills up, but there are different ways to solve that problem without extra friction.

I suspect there have been some initiatives to increase AOV, but I’d like to go further and make prices fit buyers with local Robux pricing in more countries, more plans for Plus, and auto top-up for adults who want it.

Last year, 44% of Roblox’s revenue came through Apple’s and Google’s stores (29% Apple, 15% Google), at fees of up to 30%. I’d push further with web-only bundles and by making web the default place to buy Plus where store rules allow, since store fees on a subscription repeat every month. You can flip this on its head too by acquiring customers with lower-priced Plus bundles on the app store, then funneling them to the web to upgrade their tier. This area is a product problem to me, and I believe there are a handful of ways to solve it.

To bring processing rates down, Roblox needs real alternatives to its current processors and a way to move volume between them. I’d run an RFP on interchange-plus pricing, route debit to lower-cost networks, and negotiate incentive deals with processors and the card networks. Those talks go better once the orchestration layer lets Roblox shift a country, card range, or payment method to whichever route performs best.

Collecting money internationally costs more than collecting it at home, because of cross-border acquiring, FX, and correspondent banks. Local acquiring, local-currency pricing, and steering buyers to cheaper bank rails like Pix bring that cost down as international payers grow. The orchestration layer should compare each route on all-in cost and approval rate, since the headline processor fee leaves most of that out.

Paying creators costs money too. Roblox funds the Wallet partner every business day, converts currency as Wallet goes international, and pays for the rails that carry each payout. Roblox doesn’t publish that cost, so this lever is a target in basis points on creator payouts. I’d fund the partner just in time and in the right currency, and pay out over local rails in each new market.

In some geographies, stablecoins could be a cheaper option, and I’d test them only where they beat the bank rail Roblox would otherwise use. They still carry network fees, mint and redemption costs, off-ramp costs, compliance work, and local rules, so the case has to be made one geography at a time. Creators would still see dollars, pesos, or euros, and the payout router would pick a stablecoin rail underneath only when it’s cheaper or faster.

Two cases look worth testing. One is cross-border settlement between Roblox entities and partners, where dollar settlement that runs around the clock can cut correspondent-bank hops and cash stuck in transit. The other is opt-in USDC payouts for creators who already want them, in markets where that beats the bank route on cost and speed. Airwallex, Roblox’s US Wallet partner, already offers USDC payouts in early access, so this could be one more endpoint in the same payout router.

I suspect many disputes on Roblox come from families not expecting their child to make a purchase, rather than from stolen cards. I’d try to build radar functionality where we can refund early when a dispute alert comes in, fight the disputes we should win with Visa’s Compelling Evidence rules, and use spend limits that take the buyer’s age into account. I’d agree on a loss budget in basis points with Finance and Risk, then push approvals as high as that budget allows, instead of debating rules one at a time. (I suspect many of these options are built, but we can get creative with chargeback products for parents. For example, could we give free Robux to their kids’ accounts if they avoid disputing? That would drive organic network growth because parents would adopt the Wallet control products we build.)

It looks like Wallet will be an opt-in product when invitations go out in December, and since DevEx ends in the US in mid-2027, taking Wallet from zero users to every eligible creator will take time and could delay the revenue upside we’re expecting. I’d love to build a handful of small product features across the ecosystem to drive awareness and adoption.

Once earnings land in Wallet, they belong to the creator. Each dollar they spend on ads, or on Robux at the DevEx rate for things like publishing fees, comes back to Roblox as bookings instead of going to a bank account or a card. I assume 30% of Wallet volume comes back this way, which is a guess until Wallet has real data, and the live model on the right counts it as new bookings. (Move the Wallet adoption lever above to see this one change.)

If Wallet becomes where creators run their business, more of the $1.7 billion Roblox pays creators each year comes back. This is the most enticing Wallet lever, since a portion of the money moving through Wallet becomes new revenue. Creators can already pay for ads from Wallet, and letting them set aside part of each day’s payout for ads automatically would make reinvesting the default.

Wallet can become a business of its own by earning interchange on the card, FX on local-currency payouts, and a share of yield with our partners, each of which earns a little on every dollar that flows through. Wallet’s pricing isn’t public, so this lever is a target, set in basis points on the creator payouts that run through Wallet. The goal is a revenue line that pays for the Wallet team without pricing that makes creators feel nickel-and-dimed. (Once again, I’d hold off on lending.)