Instant messaging is old, it predates the Internet, first appearing on multi-user operating systems like Compatible Time-Sharing System (CTSS) and Multiplexed Information and Computing Service (Multics) in the mid-1960s, its older than me! During the bulletin board system (BBS) phenomenon that peaked during the 1980s, some systems incorporated chat features that were similar to instant messaging, e.g. Freelancin’ Roundtable. Also an important note from history is that many popular internet services end up being fads. In the latter half of the 1980s and into the early 1990s, the Quantum Link online service for Commodore 64 computers (that was my first computer, followed by the Atari ST) offered user-to-user messages between concurrently connected customers, which was called On-Line Messages (OLM), and later FlashMail. Quantum Link later became America Online and made AOL Instant Messenger (AIM).
Modern, Internet-wide, GUI-based messaging clients (Graphical User Interface), began to take off in the mid-1990s with PowWow, ICQ, and AOL Instant Messenger. In 2000, an open source application and open standards-based protocol called Jabber was launched. The protocol was standardized under the name Extensible Messaging and Presence Protocol (XMPP). XMPP servers could act as gateways to other IM protocols, reducing the need to run multiple clients. Multi-protocol clients can use any of the popular IM protocols by using additional local libraries for each protocol. But it failed to achieve wide adoption like SMTP (email). Lotus Sametime implemented XMPP in 2007, but its business model is different to most free web services, Gtalk does use XMPP but its servers do not federate with XMPP.
Some of the benefits of XMPP are:
- XMPP provides authentication;
- XMPP provides efficient real-time two-way connectivity
- Any application can use XMPP for communication. It is more secure and efficient for clients than if each application used, say, TCP sockets; and XMPP is scalable.
- XMPP extensions can implement various services, such as the initiation of VoIP, location broadcasting, information exchange via web services, etc.
Most IM networks remained silos, as their business model is about locking-in customers, they made changes to prevent them from being used by multi-network IM clients. For example, Trillian had to release many revisions and patches to allow its users to access the MSN, AOL, and Yahoo! networks. At one point I had Skype, MSN, AOL and Yahoo IM clients running on my desktop (good job it was connected to the mains) until most people in my contact lists consolidated into Skype. The use of proprietary protocols has meant that many instant messaging networks are incompatible and users are unable to reach friends and contacts on other networks. This has cost the instant messaging dearly; as we see today MSN Messenger, Yahoo IM, AIM (AOL IM) are all been in decline as IM became embedded in other services. Skype set the bar in enabling a rich communication service beyond IM, unfortunately it’s not kept up with the market as it was bought and sold and now providers like Tango, Whatsapp, Viber, etc. are offering a richer experience with an eye on being bought by Facebook for $1B+. IM capabilities are also becoming embedded in browsers with WebRTC, as discussed in this previous weblog article
So will today’s OTT providers suffer the same fate as the IM clients of being a passing fad because their model fundamentally aims to lock-in customers to their service which stops them becoming any more than a ‘nice to have’, which then withers on the vine as they simply do not enable successful communications beyond their silo? That’s what the market is going to decide. Even if Tango is bought by Facebook, Facebook is not a pervasive network, for reasons of customer behavior and market penetration. Facebook’s model, like that of the OTT providers is about lock-in and creating a ‘private internet where the users are the product on sale.’ Hence why Open Social has never taken off beyond enterprise applications (where the business model is different.)
Joyn has the potential to offer a more successful communications experience nationally as it works across all operators with an enriched communication experience similar to the OTT providers. In the limit a customer’s preferred service will be the one that just works and happens to be bundled with their mobile ISP (Internet Service Provider) bill. Yes I did say that, mobile, like fixed has, is moving towards the ISP model. And telco boards need to get serious about their Joyn strategy as discussed in the IMS World Forum summary rather than playing around in the hope the world remains the same. This is the fundamental advantage of telcos, their business model is different to OTT, they’re focused on the customers needs as they take their money, which means cross-operator services initially at a national level are essential to remain relevant.
Just as Trillian appeared in the IM scenario, its possible an OTT OTT provider could try to aggregate across many communication clients, offering a “lowest possible cost for the greatest chance of communications success.” Smartphone’s communication experience does that today in giving you a range of communication options when you select a contact. Smartphone platforms are happy to implement whatever gives the people buying their phones a better experience, while the OTT providers will see the OTT OTT feature as disaggregating them from their users, so will likely copy what the IM platforms did in restricting the experience though perhaps not blocking it completely.
Its become fashionable to use OTT on lots of slideware to justify whatever message is being given, but please look back at history to see that OTT has its challenges as well because free means silos, and customers need things to just work. There will be continued substitution of revenue as discussed in the end of year weblog article and in the services domain report. But things are not black and white (I’ve warned previously on the plague of black of white thinking), its shades of grey, but the grey is getting darker through continued inaction, reacting not thinking, and lack of commitment.
Just one correction, GTalk does open federation with other XMPP services. AIM has an federated-on-demand-(and-I-suppose-payment) XMPP federation gateway.
Interesting historical view, but I disagree in principle.
There is huge value in silos & fragmentation in messaging *for some use-cases*. There are some instances in which you want lowest-common denominator capabilities that “just work”, but there are plenty of other cases where that is either not valuable or actively undesirable.
For example, one of the reasons that teenagers in many countries adopted BBM was because of its perceived exclusivity – it was invite-only, and only the “cool kids” had BlackBerries.
Another example is Facebook’s event invitations, which are feature-rich. I’ll send 50 invites to a birthday via FB, and maybe 5 via SMS or email to those who abstain from using it – and probably forget to send them maps or updates.
The question which is unclear in my mind is what proportion of messaging value is concentrated in open vs. closed communities. My perception is that it is shifting towards the latter, and that in many cases forced interoperability will actually destroy value if it is “imposed”.
Very interesting analysis. Firstly, I would say that XMPP is great, and it could be federates.
Once said that, I strongly believe in RCS-e/Joyn as it improves currently communication experiences that everyone’s have “internalized”, call and SMS and not so successful MMS, now the call can be enriched with video and sms improved with chat & file transfer respectively.
I do not see people would like to have a lot of different services to contact to different groups (to complicate and most sure possible overlaps) but having a common way to reach all.
Silos are not the best options and even if Joyn and open XMPP systems could interoperate (easy and already solved) telcos will add value on top of their alternative and demonstrate that it is not about exclusive clubs.
Thanks so much for making a comment.
Indonesia is a good example of where BBM is used in the use case you describe. Having a cool device is definitely an attraction for many customers, as is bypassing the SMS charges (especially for international SMS). The exclusivity of BBM is like any IM service in that invites must be accepted. But SMS/email/voice still works to those devices, so a BBM user can ignore but not hide. iMessage is another example of a closed community, virtually all the messages between my wife and I are via iMessage, but we do not make a decision to use iMessage, we make a decision to text and it just works that way as we both have iPhones and are both data connected (usually via WiFi).
We are living in a period of private internets, FB being an example. I was an early member of that private internet, but decided when they did their private placement that the culture of the organization was not aligned to my values, so I left that private internet, their IPO reaffirmed that culture. Increasingly I’m seeing others in developed markets refrain from using that private internet, just like Myspace, Friends Reunited, and Friendster before it. I was on all of them and do not use them anymore. Social is becoming a feature embedded in other services we use. There are many event planning services that use multimodal communications, not forcing people to use FB exclusively and avoiding the problem you raise on maps / details of the event.
The value in closed communities is clear, but messaging is just a piece of technology. The relationships and information shared between those closed community members is where the value resides for the community members, not in one of the mechanisms that is used to share information. The use of closed communities by free services is clear and continues a decades long tradition as I reviewed in this article.
An operator has customers that pay them, they must focus on the customer to survive. FB has users which are the product it sells to advertisers, it’s a different business model that only works if it’s a closed community else others start stealing those customers. Interop is never imposed, its chosen. Back in 1998 when interconnect was enabled between UK Operators O2, Orange, Vodafone and T-Mobile, it was not imposed, they did it because it was in both their and the customers’ best interests, and from there the rest is history. Email is another example of a service that just works. If it existed as a series of silos it would never have taken off. Interop should be an option, so those with business models to support that should be allowed to adopt it as it will be in their and their customers’ best interests.
Thanks again for your comment.
Thanks for this post Alan, I learned much about the history of IM.
I’d like to bring to your attention one siloed IM that continues to be hugely successful: QQ. Not only has QQ retained its crown as the dominant IM in China (700mln active accounts), but Tencent has also been able to translate that userbase into massive profitability (mostly by funneling its IM userbase to online games). It shows no signs of decline and has even added a siloed mobile IM service (WeChat) to its lineup that is growing very fast (200 million active users).
QQ is not an IM service that I like personally, but it is an example of a tremendously successful biz model. Tencent has a market cap of $60 billion, behind only Google and Amazon among Internet firms.
Good points, however, the Internet of China is a unique situation, though similar to some extent to that of DoCoMo’s imode in Japan. We saw the failure of imode when it was applied to other markets. And similarly the companies you mention in the Internet of China continue to struggle in taking their models into more competitive markets. The broader internet shows global scale wins, as long as China remains a closed model those companies can continue to thrive within that market, but will struggle outside of that market.