Recent news (well, still speculations) of Microsoft dropping the Live messenger in favor of Skype made me think… Instant messaging market was a fairly crowded space invaded by many proprietary solutions and one widely adopted open standard. For a long time, it seemed like proprietary solutions are winning the war, but in the end (probably due to failure of big players to monetize their services through advertising models) the last man standing will be the open XMPP standard. What the heck happened?
The dark (and fun) ages of Instant Messaging
Rewind to 1996. Although IRC was arguably the first instant messaging network for some time by then, it kept being the media of communication for nerds and geeks. Small Israeli startup Mirabilis introduced something new: one-on-one chat solution with, from the user’s perspective, zero setup, zero maintainability, no cost service. Cleverly branded product+service combo, ICQ (spelled “I seek you“) kept its domination over the newly formed market for a while. Not even the newborn issues (which in case of ICQ lasted for years) were enough to scare the users away. Besides ups and downs, their fall began after the AOL acquired it.
Although nobody knew how to monetize such service, big players rushed in. Microsoft introduced, and too often rebranded, their instant messaging service we know today under the name of Live Messenger. Yahoo! brought to life its own Yahoo! Messenger, and so on…
I guess I wouldn’t lie to you if I said that Microsoft’s service took over the market as ICQ’s influence melted. But there were niche markets where other services ruled.
For example, every big nation which wants to distance from what they think of as US-based internet has its own popular instant messaging platform. Chinese one is, for example, QQ.
Also, I have worked full time for years for a US-based company which was, least to be said, unfavorable to Microsoft products and services. Hence in the beginning we have used Yahoo’s IM service.
Also notable, many ISPs were introducing their own instant messaging services, most often just re-branded popular (or less popular) products made by other companies.
Alas, there was a choice for everybody. Nobody was earning money on instant messaging services, but everyone felt it has to provide one as an added value. Such a diverse market had one serious flaw: there was no interoperability between networks. There wasn’t even (besides IRC) any open standard for instant messaging, or even an instant messaging interoperability.
So, in 1999., XMPP is introduced. Fully open standard, soon to bring fully open implementation of cross platform server daemon. But it didn’t catch up for a while. XMPP also brought something important – it standardized the federation between services. Long story short – federation makes inter-server connections possible, while keeping communication of users inside any server private. For example, if your company (Company A) runs its XMPP service, and your client is Company B, which also runs its own server – you can use same instant messaging client, configured for the same connection to your (Company A) server only and still communicate with your clients in Company B. If both servers (Company A and Company B) allow federation between themselves.
There were couple of problems. To start with one: each of proprietary services brought its own additional bells and whistles (for example, I can safely assume that legendary ICQ feature for searching users in the users’ database targetting ones that are currently online and being able to start chatting with random people created lot of online and real life friends). Also, voice and video chats started becoming must-have feature and open standards had yet to catch up on that field. And changing the instant messaging network of choice was harder than you can imagine (if you’ve switched, you were leaving all your friends behind on your previous network).
Proprietary instant messaging vendors (at lease some of them) also started making solutions for corporations. I remember setting up ICQ’s own corporate solution inside the company which IT department I’ve managed in 1999/2000.
Competition heats up
The end of the 1990s also brought one new trend – multiprotocol IM clients. Those allowed people to become more promiscuous as instant messaging users, so many people started opening accounts on all available instant messaging networks.
You would assume big players would accept such a trend, because it brought them piles of new users and everyone was everywhere. But no. Problem was, probably, in revenue. Although nobody still knew how to monetize instant messaging services, as long as users stuck to official IM clients there was a chance to push any kind of content towards users. With 3rd party clients, specially those which supported more protocols (so they could also strip down protocol support to bare minimum of features, and often did so due to practical reasons) service provider had no means to push ads through banners in your instant messaging chat windows. It was advertisement approach which didn’t start to pay off, but no one had better idea of how to generate income on such services, so it was better than nothing.
In that light, Microsoft changed their protocol specification more than often. It was also leaving their legacy applications users behind, but Microsoft had never had problems with that. Yahoo! followed that practice although less often. Result was that 3rd party multiprotocol clients were often kicked out of proprietary networks for weeks or even months until changes to the protocol support would be implemented.
XMPP was implemented widely, but it was practically invisible to users of public instant messaging networks. Foundation behind the XMPP provided “central instant messaging service”, but the client setup was a hassle for regular Joe users and the lack of “official client” was obviously keeping them off also.
XMPP did catch up in corporate oriented IM products. Cisco, for example, provides Cisco Unified Presence product based on XMPP open standard.
In the following years, products like Skype and the availability of broadband internet managed to switch the focus from instant messaging to VoIP and similar services. Every company turned to this new field, either by introducing new products or by extending functionality of their existing IM clients. In that light, rumours mentioned in the beginning of this post don’t surprise me.
Instant messaging as we knew started to wear off. Not only because of Skype’s success, but also additional factor – Facebook.
Facebook changed, for better or worse, the landscape of the web (which is for most people the synonym for the internet in whole). Many companies started to neglect their own corporate web sites, and they started communicating their brand towards the customers exclusively through Facebook.
Facebook chat feature, in a way, harmed instant messaging and e-mail the most. It’s not like it had killed the serious usage of e-mail, but all those chats with our not-so-geeky friends (including their e-mail forwards of “take a look at this cute kitten!”) were swallowed by Facebook, or at least most of the rest of them moved to Skype for all their one-on-one communication purposes.
As with everything around Facebook, there is also a positive and negative side to instant messaging aspect. Positive aspect is that Facebook, as (maybe arguably) the biggest instant messaging platform today, can be used through XMPP gateway (meaning you can chat with your Facebook friends with an XMPP or multiprotocol chat client of choice). This is probably the reason why some of your most geeky friends appear online on the Facebook all or most of the time.
But a downside to this is that Facebook doesn’t allow federation with other XMPP services. In one aspect that is kind of undestandable – Facebook was never more than a walled garden. It has spread wide through the web pagesm but only in a tightly-controlled centralized way. On the other hand it simply sucks.
But it is is, in a way, the biggest XMPP network today with approximated around 1 billion of users, most of them being active chat users on a daily basis.
As an answer to the Facebook threat, Microsoft and Yahoo! decided to open federation for users between their own networks, but pitfully, they didn’t adopt open protocols and open federation concept, but they’ve just mutually federated with each other. Too little, to late.
Another big player relying on the XMPP open protocol is Google.
In 2005., Gmail users could start connecting and chatting to their friends using their usual XMPP or multiprotocol chat clients. In 2006., Google opened their instant messaging service to federation with any other XMPP service out there in the wild.
Google did what it does the best: it bundled open standards up with their succesful product, hence keeping the users in their ecosystem of services and products bundled together while in the same time not cutting off their users who want to use 3rd party clients to connect to the service.
Google’s XMPP support seems to be less than perfect, but out of the popular services many “real people” (meaning not only your colleagues from the IT companies) use – it’s offers far the best support and integration with everyone else.
For now it seems that, mostly due to lack of income generation on such services, biggest proprietary vendors on instant messaging markets (at least those targeting regular, not corporate users) are slowly turning away from the instant messaging market. Facebook dominates in general, Google provides alternative for the rest of us with a nice addition of supported federation, niche markets still hold on to their own solutions. Microsoft seems to be turning to Skype, Yahoo! has other problems to deal with, and ICQ becomes more and more irrelevant despite their latest attempt to rise back on the mobile platforms and on the desktop with the freshly redesigned client.
And, for me at least, surprisingly – XMPP/Jabber won this war. Sure enough, just 5-6 years ago it was the most unlikely winner in the battle where every dog ate every other dog. But today most of my communication with friends goes through Google Talk, followed by occasional Facebook chats from Pidgin, while I have only 2 Live Messenger and 1 active ICQ user left on my contact list.
And, frankly, that all makes me quite happy.