The eBay Commerce Engine, powered by Cloud Services, creates an ecosystem of content and web services that accelerate third party development, external platform integration, and expansion during the emergence of new players.
Do not mistake this concept as a centralization model. It most absolutely is not. Each platform remains independent with its own business and its own technology. It simply allows for greater federation of web services and easier access to content. It gives a common external interface that third parties can use. Each platform opts-in to share their content and/or services. You don't need all platforms to plug-in; they can plug-in whenever it makes sense for their business. They choose to what degree and what content and services they share.
eBay Inc. aspires to create the world's best online marketplace, bringing buyers and sellers together, creating opportunities, to make a difference. And eBay has some of the best online marketplaces in the world. This concept simply asks: Why isn't there a marketplace for all of eBay Inc's wonderful products and services?
The main components of this system would involve: platforms, content, services, the cloud, and consumers.
Platforms Think of this like a TV network. The network creates the content and services, but it uses local affiliate stations to broadcast it to the viewers.
Each property remains an independent business with its own platform and own technology. Moving to this model does not restrict the autonomy of any platform in any way. However, by plugging into the cloud any eBay property can benefit from the shared ecosystem. If for no other reason, sharing this infrastructure will more easily expose a platform's content and web services without the business bearing the full cost of the infrastructure.
Content These are the tv shows and movies that the network produces. Yes, yes, I know that external production companies actually sell their content to the network, but let's keep this analogy simple, shall we?
This is where many will doubt the feasibility of this effort. I contend this is absolutely possible. At the end of the day, all eBay properties have a handful of core entities: Items, Users, Messages, and Transactions. There are many, many more business-specific data objects, but these are the core.
Items - Ads (Classifieds), Listings (eBay), Products (Half), Properties (Rent), Tickets (StubHub), Search Results (Shopping). These are all still Items.
Users - this one is self-explanatory.
Messages - Email/ASQ/ABQ/Replies (Classifieds), Message Center Messages (eBay), Contacts (Half), Send to Friend/Check Availability (Rent), Tell a Friend (StubHub), none (Shopping)
Transactions - Bids/Leads (Classifieds), Bids/BIN (eBay), Buy (Half), Check Availability (Rent), Buy (StubHub), Click Out (Shopping)
Even PayPal fits this model: Payment (items), Users (users), Notifications (messages), Transactions (transactions).
Sure, each of these platforms has their own attributes for each of these entities (an Ad on Marktplaats is not the same as a Property listing on Rent). You don't need to force each platform to change their schema. You just agree upon a data spec and create a wrapper/api/producer that will make this content available to the cloud. You reserve general attribute namespaces in the schema to allow each platform to be flexible and present their relevant attribute information. Does this require the consumer to know what attributes those are? Absolutely. But they would have to know that to plug directly into each platform anyway.
Services Networks have television production companies, movie divisions, and advertising departments. These are some of their services.
Let's be honest, most of the services will be specific services offered by a specific platform. And that's okay. Marktplaats may offer a license plate-based autos search for cars in the Netherlands. PayPal may offer Send Money. StubHub may allow you to check for U2 concert tickets on StubHub (and only StubHub). This is no different than what a developer/consumer has to deal with today.
Ah, so why bother? Okay, so let me ask you to tell me every service Kijiji has to offer. How about Mobile.de? What do you know about Bilbasen services? Now what if I told you you need to go to one and only one place to see a listing of all these services. Better yet, one place that has all the API documentation and tools.
Further, LoQuo will have to deal with how they throttle API users from overwhelming their systems. So will Den Bla Avis. So will Gumtree. Federating your service puts the burden of infrastructure in one place that can be scaled, hardened, and extended.
So service federation is not about making a single license plate search service that works across eBay Motors, Marktplaats Autos, Mobile.de, and Bilbasen. Who would actually use that? It's about making it easy for external developers to get at those individual services.
Now, some services could be global, multi-platform services. These would be larger core competencies like payments or search. Why does each platform need its own shipping cost calculator? Why pay for maintaining that? Does 1+1 equal something other than 2 on an Italian or a Canadian calculator? They're the same, right? Worse yet, what about a new platform that doesn't have some of these big core services? Should they have to blow a bunch of development to do essentially the same thing?
Expand it even further. What if I am a game developer in Las Vegas or a new, hip hotel in Shanghai, why wouldn't I want a plug-and-play payment service? Better yet, one that has the biggest real-time fraud prevention database in the world? I think there's some value in that.
Federating services is not about consolidating all services. It's about putting your services on display and making it easy for people to use them. And for those services that do make sense to consolidate, you win on savings and create a new revenue opportunity by selling it to others. eBay Inc's services themselves are revenue opportunities, not just the products it sells.
The Cloud The Cloud is like the fiber optic cables and affiliate tv stations.
The cloud simply distributes the content and services. It addresses operational excellence, redundancy, latency, and performance. It makes it so a business in Bangkok doesn't have to go all the way to San Jose to use eBay's best-in-the-world Product Search technology.
Consumers In our analogy, consumers are your TVs, your computers, your TiVo, your SlingBox.
iPhone. Blackberry. Android. Palm Pre. Airplane In-flight Entertainment systems. Supermarket shopping cart consoles. Firefox. Internet Explorer. Digital billboards. Cable TV set top boxes. Facebook. Virtual gaming. PSP. Nintendo DS. Wii. XBOX 360. PS3. Elevator info panels. Bus stop kiosks. Post office bulletin boards.
Where do you want eBay Inc. services? Where do you need to make a PayPal payment? What city do you want to do a local classifieds search in? Where is commerce happening?
Could a third party plug directly into each platform? Sure. But each integration becomes a one-off development cost and some very fundamental building blocks would have to be replicated by each platform. For example: storage, throttling, authentication, API gateways, and message brokering. You've also got to deal with latency, asynchronous de-coupling, and operational maintainability. This is costly and very few businesses can support this business case unless you've got the volume and revenue of eBay or PayPal. It is unlikely that Rent, Gumtree, or LoQuo could afford something like this. StubHub, Kijiji Canada, Marktplaats, and Mobile, perhaps; but it would come at a business opportunity cost that would likely be too high.
Prior efforts focused too heavily on centralization and standardization, when the first priorities needed to be access and flexibility. Lower the barriers for platforms to plugin and reduce lock-in, and you're more likely to get adoption. Once each business sees the benefit and as volume starts to force the discussion to matters of scale, then these platforms are more likely to come seeking a simple solution that can be reused. Reused, not centralized.
eBay's eBox and some components of the corporate architecture were a good start for this. However, this idea needs to be expanded to include platforms which do not use the core eBay architecture. eBox focuses too heavily on the eBay.com technology stack and data structures. For Proteus to work, it has to be inclusive. It has to be easy. It has to be fast.
Why do any of this at all? Awareness. Perception. Customers, Businesses, and Developers view eBay Inc. as a bunch of destination sites. When they need to be thinking of eBay Inc. as the world's most powerful commerce engine.
Is eBay just "that online auction place," only good for used or second-hand items? Is PayPal only a way to send money via email? Is Skype just another chat or VoIP client? Are Marktplaats.nl, Kijiji.ca, Gumtree.co.uk, Mobile.de, DBA.dk, LoQuo.es, just distant, international foreign classifieds sites?
Or...
Is eBay a commerce platform, powered by the world's best commerce engine? Is PayPal the only global, real-time financial transaction system in the world? Is Skype a distance shifting technology, bringing schoolchildren in Darfur in touch with teachers in Topeka, Kansas? Is the eBay Classifieds Group the greenest engine fueling local economies?
Imagine this, a platform that has:
- the best product search in the world
- the most extensive and unique inventory in the world
- the easiest, most secure, and only real-time payment transaction system in the world
- the most convenient way to connect via chat, voice, or video
- the strength of local classifieds, local availability and bargain inventory
- the most passionate and loyal Community of Buyers and Sellers in the world
Read more in Proteus Part Two and Proteus Part Three.
--- Why Proteus? Proteus was the Greek god of transformation, known for changing forms. He gives us the word "protean," meaning changeable, adaptable, mutable, and versatile. eBay Inc's products and services can take many forms. Hence, Proteus.
BTW, Proteus used to change his form to avoid being caught and being forced to foretell the future. Where will the future of commerce be?
conceptsecommercegadgets
