Built and evolved since our call on 10 September, from ZBD's internal competitive paper, and the Roblox/Airwallex note to the board. An outsider's first read, offered to be argued with: the state of the company, the areas that seem to matter most, the competitive position of Payouts and Accounts, and — at the end — where I could potentially help.
Confidential — contains non-public financial and deal informationEmbedded Rewards never found repeatable fit — one customer drove nearly all of it, peaking at ~$1.2m a month — and is being deprecated. The business restarts on Embedded Payouts (live since August, first customer, hardening through 2026) and Embedded Accounts (MVP at Gamescom, usable by end of Q1 2027).
They dislike Hyperwallet and like the product; ZBD is unproven. Second technical DD call in late September. Scope is one segment of the Epic Games Store as a test, not a wholesale replacement. Sponsor has changed; Gordon is driving it.
Roblox Wallets — creator accounts, same-day payouts, cards — is exactly ZBD's vision, built by Roblox on Airwallex rails. It validates the category and fires the starting gun; Epic, Sony and Valve will have seen it. Airwallex is now the most direct competitor by scale, Tilia by product shape.
The product differentiation is real. What is thin is the pipeline behind Epic, the public evidence, and a website that still speaks to mobile developers. Runway is finite.
Payouts live (drop-in screen and API), hardening through Q4. Accounts MVP shown, usable end Q1 2027. Rewards fully deprecated. The site still leads with Rewards.
Simon (CEO); Andre (CTO, consumer-product background — senior fintech engineering experience alongside him is a known, expensive need); Nicole (product; built Tilia's product before ZBD's); Gordon (CCO, twenty years PlayStation Store) and Nate (ex-Xsolla) sell; Ben opens doors; one junior BDR. Compliance in the Netherlands, tech and product on the US West Coast, developers in Brazil.
Epic (segment test, ~50%). Fenris Creations — formerly CCP Games, EVE Online — CEO relationship, Accounts use case. Roblox lost to Airwallex. Mid-tier PC/console creator payouts, modding and UGC platforms as the broader field. Cash-gaming aggregators (Mistplay, Freecash) as a channel to test. Real-money gaming under compliance review.
Board dynamics much healthier than they were. RAINE represents the A and B rounds; Nico's fund the C. A four-year convertible-note structure is punitive to A and B holders if revenue is slow — worth naming rather than managing around. $5m of the Series C unfilled; Sony a candidate, with Sony Bank as a route into Japan.
Epic is 50/50 and Epic-sized deals take time; Gordon and Nate are the only two who close enterprise. The plan that seems most robust is "Epic and Fenris and five more accounts of that shape and a channel test", with named owners and a pipeline view the board sees regularly. Good: a pipeline the company survives if Epic slips a quarter. Worth asking whether the effective BD from the old team is recoverable.
Gordon's team has taken it from C-suite interest to a second technical DD in months, against an incumbent Epic dislikes. The point for the plan is what the deal means either way: a win on a test segment is the reference that changes every other conversation; a slip is survivable only if the rest of the pipeline exists. Good: the test segment defined, the three open items in ZBD's own paper closed (finance policy, support plan, technical depth), and a plan that holds in both outcomes.
The site was built for mobile developers, leads with Rewards case studies for a product being retired, never names the enemy, and does not say the three things that beat Airwallex: game currencies, the in-game ledger, in-game cash-out. Marketing is being cut just as the market gets contested. Good: a site a Fenris or Epic product lead recognises their problem in, with the Airwallex and Hyperwallet answers on the page. Something I could potentially help with.
Roblox gave Airwallex the reference logo. ZBD's public evidence is Rewards in mobile-casual and one unnamed live Payouts customer, so every enterprise conversation starts from a deficit. Good: a case study with numbers from the first Payouts customer, and Fenris or one mid-tier title lined up as the second.
On the Epic calls, the people who responded most were on the Unreal Engine side — the game and creator-tools people, not payments or treasury. What lands with them is their own title with ZBD's cash-out screen in it, not a product video. Good: a repeatable, half-day process to produce one before the first proper meeting. Possibly something I could help prototype.
As the products move from consumer rewards to regulated financial infrastructure, the engineering leadership need changes shape, and a full-time senior leader is expensive against a $1m burn plan. How much of the need is one-off (an Accounts ledger architecture review, a credible engineering face for diligence) and how much is ongoing? Could a consulting engagement cover two quarters while a permanent role is scoped? There is at least one person the team has met who could fit that shape. Good: a clear view of which shape fits, with a cost, before Accounts ships.
There is payout volume, a willing US banking partner and operator appetite; there is also no in-house BD expertise and a set of state licences to protect. Quantify first — volume in reach, from whom, at what margin, what the banking partner and regulators tolerate, what BD capability costs — then decide. Separately, give Rewards publishers a clean path (direct Payouts API, or the aggregator channel) so the exit doesn't become the story. Good: a one-page RMG sizing with a recommendation either way, and a customer-by-customer wind-down plan.
$2m monthly revenue and $500k monthly gross profit by December 2027, on a runway to end 2028, is a target every discussion can be measured against. The A and B holders lose under the note if revenue is slow; the C holder's position differs. Good: a one-page scorecard at every board meeting, and the incentive question on the agenda once.
$1.5m a month against revenue heading to zero is a lot of company; the fixed-cost plan takes it to $1m, which is still $12m a year for a business whose revenue restarts from nothing. The question is less "is it too high" than "what is each part for": people (and where — the Netherlands, the US West Coast, Brazil), the licences and compliance overhead that have to be carried regardless, infrastructure, Rewards-era costs that can go with Rewards, and anything that exists because the old plan needed it. A zero-based view — what would we spend if we were starting Payouts and Accounts today with the same licences — is usually more useful than trimming the current list. Good: a one-page breakdown of the $1m by purpose, each line tied to Epic, Accounts, licences or keeping the lights on, and an honest number for what is discretionary.
Everyone else moves the payout moment out of the product or makes the publisher build the experience.
Whoever pays the fee sets the minimum, and the minimum decides who ever gets paid.
Cashable game currencies, one ledger that knows where every unit came from, compliant P2P. Airwallex has none of it; Tilia has a version — that contest is on depth, and Nicole knows the lines.
Stripe's own terms bar it; nobody else treats it as a product. ZBD is licensed for real-money gaming flows.
Unity plugin and drop-in screen in days; sub-cent, instant payouts. Airwallex serves large accounts only; Tilia is the only other engine-native option.
Once Payouts is live the account already exists. Marketplaces, P2P trading and creator economies are the areas your note calls "ours to take".
Epic-shaped accounts, run as a defined test segment. Lead with onboarding pain and the threshold; enter via product or creator relations, not treasury.
Real economies, reachable CEOs, below Airwallex's interest. The segment that turns one whale into a pipeline. "You need five more of these."
Anywhere Stripe has said no. Fast decisions, quantifiable pain, natural cross-sell into Accounts.
Creators to pay, trades to make legal; too small for Airwallex, often on Stripe, sometimes Tilia or Tipalti.
The Take-Two / Unity / Ubisoft / Zynga / Voodoo tier that has hit the Stripe-branded onboarding wall or can't pass the fee through.
Mistplay, Freecash and two or three peers: sell them Payouts and take the flow. A dated kill criterion before it starts.
Where the first question is 1042-S handling or 200-country coverage, ZBD loses today. Take the segment, not the rail.
No one in the payouts set offers it; Airwallex offers the bank-account half without the economy half. Bigger deal, deeper stickiness, longer cycle — won on the back of a Payouts relationship that already works. Usable by end of Q1 2027 is the date the pipeline has to be built against.
A community-driven spaceship MMORPG with a thriving player economy across 7,000 star systems: war, politics, piracy, trading, exploration. Its developer, CCP Games, has rebranded as Fenris Creations after splitting from Pearl Abyss. You know the CEO. Every part of Embedded Accounts already exists there in virtual form — a currency, markets, player-to-player trade, contracts. Making a slice of it cashable, compliantly, is the cleanest possible proof that "the money layer for games" is real.
Close Epic's three open items in writing. Get the Airwallex answer into the team's hands before the next prospect opens with "we saw Roblox". Start the message architecture. Complete the Zopa sequencing so ZBD can announce.
The scoreboard, a monthly pipeline view, the Rewards wind-down plan, the Fenris pilot scoped, the mid-tier target list, the note-structure tension named.
MSA architecture, SLAs, exclusivity and roadmap allocation agreed before signature. Senior engineering role scoped. Aggregator channel tested with a kill date. Sony conversation alongside.
Epic go-live and payee migration off Hyperwallet — where displacement deals actually fail. Fenris or a trading-first pilot on Accounts. Three to five mid-tier Payouts wins.
Accounts references at enterprise and mid-tier; Japan via Sony brought to the board with licensing costed. Runway ends in 2028: either the number is real and the company is close to self-funding, or the next raise happens in H1 2028 on the back of it.
Because Roblox built the creator experience itself on plain rails, and has the engineering to do it. Airwallex gives a creator a business account and a card; it has no virtual-currency conversion, no in-game ledger, no player-to-player economy, and it doesn't serve studios below its commercial threshold. If you have Roblox's build budget and only need to pay creators as businesses, it's a fine choice. If you want the finished in-game flow and an economy every player can take part in — now — that is what ZBD is for.
Retire the Rewards numbers ("+1,215% D30", "+355%", "+45% ARPDAU") from the enterprise narrative — the product is being deprecated. The Hyperwallet cost estimate for Epic (4–6%, ~$21M) is directional. "Crypto company?" — ZBD is a licensed payment institution using Lightning as a settlement rail the way Visa uses VisaNet; the site still leads with Bitcoin in places, which is a positioning decision rather than a product limitation.
Three claims, the enemy named, the Airwallex and Hyperwallet answers on the page; possibly a working prototype the team can ship rather than brief an agency.
A reusable mock-up of a named prospect's title with the drop-in cash-out screen and an Accounts balance in it — Epic-shaped, or built on EVE for Fenris.
A one-page battlecard and a paragraph for the site, for the next Epic, Sony or Valve conversation that opens with "we saw Roblox".
Gordon has the deal well in hand. If a second view helps at some point: where the evaluation stands, what the sponsor change and the Roblox news do to Epic's options, and the risks from outside.
Help think through hire versus fractional versus consulting, and if useful introduce the engineer the team has met and a payments-infrastructure architect for a second opinion on the Accounts ledger.
Twenty specific, answerable questions across GTM, product, team, compliance and capital, each with what "good" looks like. The nine areas above are the headline version.
If useful, sit with you or whoever owns the numbers on the $1m breakdown and the zero-based version of it — where the money goes by purpose, what is licence-driven and immovable, and what is discretionary. A read, not a plan.