January 06, 2007

Skype drops the payments bombshell

Canny financial cryptographers will spot the bombshell in the first and last comments of the article mentioned earlier on Skype. Read those paras first (look for "payment") and come back, as the rest won't make sense.

Skype are adding payments. I may as well now reveal that we brave cryptoplumbers at Systemics built this over the period 2001-2004. (Ooops! I already revealed it.) It took about 3 years to do, as the 6 week "summer edition" by Edwin Woudt proved a need to re-jig many fundamental parts of the SOX & Ricardo architectures. The rewrite (started by Jeroen van Gelderen and Edwin, finished by yours truly) worked, and the chat part worked so well that I can predict that it would be addictive if we ever fielded it. (Why it is not fielded is your mystery and our loss.)

So what's special about this? Someone else figured it out: Skype. They are in the process of implementing Payments over their infrastructure, which means it is no longer necessary for me to keep quiet about this innovation. We've been overtaken, so I may as well reveal all.

FTR, we implemented chat over a payments system, whereas Skype are implementing payments over a chat system. If constructed as a core FC application, the result is indistinguishable, for reasons that I am slowly writing down in many parallel documents. I continue to develop the underlying infrastructure in my spare time, as well as document the core concepts, and you have already read many of the facets on these pages ... but the absence of the whole story should give you a hint as to why this is not a fielded app.

The big picture is this: integrated chat & payments is huge. Immense. It has the potential to be if not the next killer app, certainly the next killer integration. I can show this by analogue: in your last 10 payments, how many messages did you send to your counterparties? Chances are, at least 100. Ergo, the message that carried the payment is the least of the protocol known as trade.

Trade is a chat application (with a payment message thrown in somewhere around the end).

The big question for those who appreciate this and are at this minute going long on Skype is ... can they do it? Here's my answer: Yes, if they take small baby steps, then they may avoid the many bear traps of financial cryptography. That is, they at least have the track record for doing this in lower layer terms, and if they don't rush it, they'll pick up the hard higher layer lessons in time.

Can others do it? Not a chance. Skype have the field fully open. Google, AIM, Jabber, and that MS thing (I don't use it so don't know what it is called) are so unsuited to the financial cryptography needs of chat plus payments that they haven't a chance. To be honest, it will probably require dramatic changes in Skype's architecture too -- the difference is that they have *enough* in place to show they can pick up the rest, and survive the transition. The others have no chance, IMO.

Even though I failed to deploy this and am competitively annoyed, it is a joy to see these things evolve and for others to pick up the baton. Go skype! What others should bear in mind is that this is an A-grade FatBoy BombShell, for reasons that can't really be explained in a simple blog post, but will be seen when and if Skype deploys it.

If they falter, we have to wait for the next time.

Posted by iang at January 6, 2007 10:55 PM | TrackBack

> The big picture is this: integrated chat & payments is huge. Immense.
> > It has the potential to be if not the next killer app, certainly the
> > next killer integration.

I don't see it happening, at least not with skype due to the fact ebay/paypal own them and ebay has a shocking track record and any one doing serious payments don't want to run the risk of ebay/paypal locking their account.

There is possibly a number of other reasons why this won't take off, not like web payments, since we already have large issues with phishing and fraud on IM networks I don't think banks will be keen on this idea.

Posted by: Duane at January 6, 2007 05:05 PM

Hi Ian

Had to happen.

Instant money messages.

I've been saying this for at least four years. You've been doing it for longer.

Posted by: Chris at January 6, 2007 05:06 PM

Also see: http://www.connectotel.com/skype/skypepaymentapi.pdf

Posted by: sizeof at January 6, 2007 05:31 PM

Marcus, that's an interesting proposal. Is there a sense that we already understand what model of payment Skype are aiming to do?

Posted by: Iang at January 7, 2007 08:30 AM

Nah, I am still hoping to have a working mobile messaging infrastructure suitable for FC (not just payments) rolled out before Skype does it. The mobile market is far bigger and far more important than PCs. And besides, cellphones are more suited for payment as a platform.

Posted by: Daniel A. Nagy at January 8, 2007 04:58 AM

Hmmm... Daniel, questions abound ... "Nah" to what?

Can you inform your comment that "mobile is far bigger / more important than PCs" with some metrics?

Messaging is suitable for FC not just payments: agreed :)
Cellphones are more suited for payment than PCs: agreed :)

Posted by: Iang (an early too-enthusiast) at January 8, 2007 10:07 AM

Surely the cryptographic requirements for the protection of communication over Skype and the authentication of the origin and destination of a call are different from the requirements of a payment system. So Skype's existing cryptogaphic -publicly apparent- expertise doesn't seem to be relevant with respect to payments.

The IM infrastructure Skype brings in is basically a "ubiquitous" pervasive peer-to-peer messaging environment with a huge and loyal installed base. Whether the trust in Skype readily will transpose to trust in Skype as a payment system provider is something different. May be Skype has investigated this potential in a marketing research program.

Of course, the pervasive infrastructure Skype provides; facilitating user registration, user authentication, application and update deployment is an important asset, neatly poised to service any payment infrastructure not based on the -IMO unwarranted- trust in SSL.

But then I loose sight of any relevance of Skype playing an expert role in payments. But then again is has been more 10 years ago when the electronic and internet payments where introduced as an innovation. May be now the market is ready, just.

In other words, I think that eBay is betting on the wrong horse if the only reason she bought Skype as a step into the direction of providing a payment system provider. On the other hand the incentives are great. The low value high volume transactions in the eBay market make it attractive to eliminate the intermediary CC fees. But hey! Is PayPal not owned by eBay.

May be the marriage of the administrative infrastructure provided by PayPal and the technical infrastructure provided by Skype can generate an interesting business case.

But I leave the analysis of that business case to you, as I don't hold an MBA ;-)


Posted by: Twan at January 8, 2007 11:40 AM

> Hmmm... questions abound ... Nah to what?

I disagree that Skype will be an unchallenged monopoly in the payment-in-messaging market. I think, there is still room for competition, especially taking into account the fact that any one such payment system actually helps all the others. Besides, WebMoney has had its messaging system for invoicing, payment, dispute resolution and, well, (authenticated, reliable) messaging.

> Can you inform your comment that "mobile is far bigger / more important than PCs" with some metrics?

According to some 2005 study (feel free to google for the original source) Western Europe will reach 100% cellphone penetration early this year. It was widely circulated and quoted in the industry. The PC penetration is nowhere near that (it's about 70% and seems to have reached saturation).

In the US the two are comparable, but that's just because the retarded
cellular infrastructure in North America.

It is more important, because mobile telephony is just the perfect "commoditized service", with a demand far beyond that of skype-out. Skype cannot charge for in-network traffic (because the market price is 0; if they start charging, their user base will be gone in the blink of an eye), so that doesn't count.

Posted by: Daniel A. Nagy at January 9, 2007 04:31 AM

Its a big task, what kind of far off date does anyone see for an actual product testing?

Posted by: Mark Herpel at February 9, 2007 08:38 AM
Post a comment

Remember personal info?

Hit preview to see your comment as it would be displayed.