Rich Communications Suite: Really Considered Significantly Obsolete
It was a curious Mobile World Congress last week; half Telco 2.0 triumph, with things like the OneAPI project, O2 Litmus, and a truly impressive focus on developer communities, and half a harking back to the days when IMS was the solution, whatever the problem might be.
Take the GSMA's 'Rich Communications Suite' (RCS). We've discussed the imperatives for voice telephony recently here. So, we're at a loss as to the relevance of RCS to the market, as one of our analysts vociferously describes below.
But we'd value an open discussion with readers who support the initiative. Do read the analysis in the rest of this article, and tell us what you think via the 'comments' function on this blog...
Rich Communications Suite - sounds impressive, doesn't it? You might be surprised to find that in fact, the feature set and the look-and-feel are identical to essentially any IM client you may have used since 1998 or thereabouts. This is because so far, that's it - RCS mostly consists of adding presence-and-availability to your mobile address book. Which is nice, but as the French say, it won't invent gunpowder.
The vendors involved - and it's practically all vendors, usually a bad sign - are very proud of the efforts they've made to make sure it's fully interoperable. Dozens of testfests have been carried out. Trials with small groups of users. All that remains is to deploy the thing and put it on the market. Does this remind you of any other recent mobile technology? Ah, you're ahead of me. IMS. And indeed, all the RCS applications they showed off were connecting to their SIP applications servers through an IMS core network. Just to prove that it really is IMS, one of the reference applications is the inevitable video-sharing.
This raises a couple of issues. To start with, SIP is not very new. It's also a proper, tried, IETF-standardised technology. Interoperability should be about as much an issue as it is with the Internet itself. You follow the relevant RFCs and things work. So, this is a sign that one of the big interoperability issues with IMS may still be out there - the fact that 3GPP chose to fork SIP and use its own non-standard standard. Similarly, XMPP is, well, a standard. And neither technology cares very much what kind of device is involved - XMPP even provides for multiple clients logged into the same server with the same user ID at the same time, routing messages based on user-defined priority. And its notion of "transports" allows you to hook up all your other services and see all your contacts in one place, which happens to be the RCS unique selling point.
But perhaps the point is to spread the service beyond the power users? Well, some of the people involved described the target audience as "early adopters, young pioneers, young materialists", which when you parse the marketspeak means exactly the people who already use IM. So it's not a good sign that neither Apple nor RIM was present. Social graph applications stand or fall on user adoption, and iPhone and BlackBerry users are strongly represented among the user elite who tend to have huge contacts networks. Neither were the Joint Innovation Lab carriers - Vodafone, China Mobile, and Softbank - around. Although the client in the demonstration was on a Nokia device, and NSN people were in evidence, we didn't notice anyone from the Symbian/software side of the Nokiasphere either.
However, others seemed to think the point of RCS was to bring presence-and-availability and IM to the bulk market, and that IM was incredibly complicated and daunting. But it can't be targeted both at the power users and at the mass market - you have to choose.
Also, we're concerned about the business model, or lack of one. Insofar as the tawdry question of commerce was discussed at all, it was suggested that it would lead to "more total communication", which apparently must be good. Ask a UK DSL operator how more total iPlayer worked out. More traffic is certain to lead to one thing - more cost. Whether it leads to more revenue, and whether this exceeds the additional cost, is up to your own creativity, product leadership, and customer intimacy. In an environment of falling prices, just pushing volume is not enough.
It's unlikely that IM traffic would ever be enough to harm operator business models, but the RCS team are awfully keen on video - live, streaming, full motion video on shared-medium wireless networks. Unless the operators can invent a meaningful business model for this, the impact of successful RCS would be indistinguishable from the impact of a hit OTT video app, which is what IMS was meant to save them from. Watch out! Also, what happens when a prepaid RCS customer runs out of cash? Do they just vanish off everyone else's radar screen, or is there to be a Smile-like model where presence, status updates, and other pre-call activity is free, but voice or video is chargeable?
Further, essentially all RCS's current and even suggested future features are already well served by competitors, many of whom have the inconvenient property of being free. Are we really expected to be impressed by "sharing photos" in 2009? When every Nokia cameraphone ships with a client designed to upload photos to Flickr (or anything else you may like)? Yes, RCS might integrate them in one place - but it's worth remembering that "one application to rule them all" has been surprisingly unsuccessful on the Web. Specialisation, according to Robert Heinlein, is for insects. But there are a hell of a lot of insects in the world. Bees are a traditional image of bourgeois trading wealth, civic pride, and work ethic for a reason. And come to think of it, bees are nothing if not devoted to specialisation.
Another problem is that RCS is reliant on someone not solving the same problems in the client. It's widely assumed that any download or installation is a massive turn-off; but the industry is frantically, and rightly, launching app stores and building developer platforms as fast as it can.
Nokia, for example, already has a unified XMPP-based address book and messaging/sociability app with Nokia Maps integration in beta, has signed up to deploy native Skype clients on the N-series, and also has both a native SIP client and a Gizmo implementation for S60. (Gizmo, like the XMPP/Jabber boys, has the ability to bridge other IM and VoIP networks and act as a single contacts book.) So RCS is relying on the failure of operators' efforts to encourage users to download and use new applications. It is a fundamental rule of success in all information technologies that you shouldn't bet against applications development - the evolutionary cycle is much faster than it is for all the other elements of the system.
And, as they say, if you prepare to fail you'll get what you prepared for. There is also the big IMS question, which is as always simply "why?" You have an endpoint with an IP address. You have an application server with an IP address to which it connects using SIP. Why do we need an IMS between the two? The force of these objections is well shown by the absence of IMS in general at this MWC - nobody is advertising it, talking about it, or indeed doing anything with it. We suspect its sudden rebranding as RCS reflects a push by certain vendors whose heavy investment in IMS begins to look like a waste.
It's also worth noting that essentially all the operator support for RCS comes from France and, to a lesser extent, Italy. This may reflect a lower penetration of Web-based social network apps there, which could provide a glimmer of a chance - but it may also imply that RCS will be at best a technology that succeeds in one market but doesn't travel, like i-mode rather than GSM.