The two approaches in one paragraph each
With direct integration, you agree a contract with a provider and build against that provider's own interface: its way to list games, to launch them and to talk to your wallet. Each additional provider brings its own contract and its own wallet interface, and your team builds and runs each of them separately.
With an aggregator, you build against one interface that covers the games of many providers. There is one way to list games, one way to open a game and one wallet interface. The aggregator deals with the provider side and presents a common contract to you.
What you build
Direct integration means repeating the same set of jobs for every provider: reading its catalog, launching its games, answering its wallet calls, handling its errors and testing the whole path. The jobs are the same; the details differ each time.
With WGC you build these once. The games request returns the games issued to your account, 200 per page, each with a publicId that identifies the game in your catalog and a universalId that is the same for the same game across the catalog. After the first full download you ask only for changes. A session request needs the player id, the currency, the mode and one game number, and returns a launch link. For the wallet you build one Callback handler with three commands, whichever studio made the game.
The wallet
The wallet is where the two approaches differ most. In a seamless wallet model the player's money stays on your platform, and the other side calls you to read a balance, record a bet and its win, or cancel a bet. With direct integration, each provider defines its own version of those calls, so your platform ends up with a separate wallet contract per provider.
With WGC there is one Callback for every game. Each call is signed with HMAC-SHA256 and your account secret. A bet is recorded once per transaction id, so a repeated call does not debit the player twice. WGC waits three seconds for your answer. If a studio sends the bet and the result separately, WGC joins them into one round, so your handler sees one shape of call. When your answer does not arrive in time, the bet is refused to the studio and WGC later asks you to cancel it with a rollback, repeated on a schedule for up to 30 days.
An aggregator does not remove the need for a wallet handler. You still build and run one, and it must be correct about signatures, repeated calls and cancellations. What changes is that you build it once.
The catalog and who decides what you see
Directly, you see what each provider offers and you agree access with each provider. Through an aggregator, you see the catalog of its studios, and the games issued to your account are decided by your manager, studio by studio. The catalog of WGC lists 184 studios, and a studio's entry does not mean that all of its games are open to every client.
One practical benefit of a common identifier is recognition: the universalId is the same for the same game across the catalog, so your own database can key on a number that does not change.
What an aggregator does not erase
An aggregator makes the interface common, but it does not make every game identical. Some differences stay visible, and an honest comparison should say so.
- Game settings. For a launch you can pass an RTP between 75 and 96 and a bet range, but only games that accept them use them. Catalog flags tell you which games do, and a game that does not accept the values still opens.
- Restrictions. Each game lists its restricted countries and the currencies in which it does not open. A launch from a known restricted country is refused. If the country cannot be determined, the game opens, so your platform keeps its own geographic check.
- Refusals. When a studio refuses a launch, you receive one of a fixed set of WGC reasons, not the studio's own code or text. That is simpler to handle in one place, and it also means less detail than a direct connection might give you.
Reporting and onboarding
With direct integration, you rely on what each provider reports, one provider at a time. Through WGC the back office has one set of reports across studios: turnover, wins, GGR, RTP and commission for a period, by round, session, player and game, with CSV export.
Onboarding tools also work once. Test keys open a test game that makes real calls to your Callback URL without money. The back office has an integration check with a live log of every call and an integration test with 14 checks. Ask WGC, the assistant in the back office, answers from the documentation and, in integration mode, reads your own recent calls to explain why they fail.
When direct integration can make sense
Direct integration can be the simpler path when you need the content of a single provider, or a feature of one studio that no common interface exposes. In that case one contract and one wallet handler may be all you need, and an extra layer adds little.
When an aggregator fits
An aggregator fits when you want the content of many studios and do not want to build, test and run a separate wallet contract for each. It also fits when the same small team has to keep catalog, launch and wallet working, because those jobs exist once instead of once per provider. If you are weighing candidates, the article on how to choose a game aggregator lists the criteria to check.
Questions to settle before you choose
- How many wallet contracts would your team have to build and maintain?
- Who decides which games your account sees?
- What does a refusal look like when it reaches your platform?
- Can you test the whole path without real money?
- Which reports will you have across all content?
For the interface itself, see the Casino API overview, the game aggregator page, the provider catalog and the contract of the seamless wallet.