2011-03-06

Quo Vadis?

Someone recently posed me the following question in an email:
Consider a leading communcation provider in 2014. (Including what is now known as "telcos" if such a thing still exists.) What are the communication functions such a provider will provide? How would you as a user expect the provider to help you with your
daily communication needs?
Rather than attempt to answer it directly, which may or may not be possible, I thought that it might be helpful to understand where we are and what the context of this question really is.  The following blog post is a somewhat edited version of the response I gave.



The fundamental problems of communication technologies today is fragmentation and low quality.

Fragmentation

Fragmentation arises from the fact that we have a gazillion services for communicating and there is a rather large space of modalities for communicating (public/private, synchronous/asynchronous, voice/video/text, etc.)  Many of these communication facilities are tied to a particular service provider (Skype, Facebook, Google, Twitter) while some are interoperable (telephony, email, Jabber etc).

If you look closely at your social graph you will notice that in order to reach everyone you know effectively, you will most likely have to use half a dozen different communication systems.  Some people respond quickly to email, some prefer Twitter, others live on Facebook, some still have land-lines to their house and know nothing of this newfangled intertube stuff.

This is a challenge.  Not only because it is inconvenient for the user, but because it is very hard to identify how one can build a one-size-fits-all communication product that reduces the amount of friction and chaos.

Low quality

The other serious issue today is low quality.   When we moved from land-lines to first generation mobile networks, we accepted more sketchy availability and a noticable drop in audio quality.  GSM was a bit more consistent in areas with good coverage, but still not great.  Skype and other VoiP-offerings can sometimes offer superior audio quality, but at the price of even worse availability: you never know if it'll work.  Facebook and Twitter offer no meaningful SLA so your message may or may not be delivered.

Standards

A way to address fragmentation is standardization.  But standardization is slow, and sometimes contrary to commercial interests.  In particular in later years actors in the market are never content with a strategy that makes them a big part of an even bigger ecosystem:  they want to own the ecosystem. and since everything is in flux these days, there are a lot of companies jockeying for dominant positions. (Perhaps the most promising may be Facebook since they have 600M users by now (or nearly 1/10 of the world's population).  With that kind of reach they have the potential to have an enormous impact in whatever they choose to do,  standards be damned:  when you are this large you can unilaterally set the standards and force everyone else to follow).

Also, there is standardization and then there is standardization.

The telco way of standardizing is dead:  a very formally driven process that takes many years to complete,  accumulates vast complexity and then leads to different players adopting different interpretations of the "standard" may have worked when we only had to deal with a market partitioned neatly into countries.  But that world no longer exists. The market is global.

There is also the "Internet way" of standardization:  RFC's ratified by the IETF.  In order to create an internet standard you need at least two implementations of which at least one must be openly available.  in other words:  it is an implementation-driven approach.  this cuts down on complexity considerably and tends to favor simpler, more compact and more implementation friendly standard.  (It is actually easier to show why this is so by looking at the antithesis of these standards.  Things like X.400, X.500 and SGML for which no fully compliant implementations exist because these standards are hopelessly complex and immensely huge).

Global market

Any communications player not thinking of everything they do in terms of a global strategy is wasting their money and time and will eventually disappear. This applies to all aspects of what telcos do today.  (I start every meeting I have with people by asking them if they have a global strategy for their project.  Almost everyone admits that they either have none, or they are "perhaps thinking about expanding to neighboring markets".  This is frightening and essentially means that money is being wasted on doing things over and over again with very little impact beyond a single market).

Pace

Technology and trends are moving faster every year.  The iPhone is about 4 years old now, and we are on the FOURTH generation of iPhone and the third or fourth generation of Android phones (how do count generations on Android?).  More has happened in the handset world in the past 4 years than in the entire history leading up to 2007.

Traditional handset vendors are struggling to keep up.  Some do a little better than others, but the big winners are not the same companies that were dominant just 4 years ago.  In particular Nokia provides a frightening example of how quickly a success can turn into a sob-story.

As for telcos:  they were caught out.  Telcos have had the privilege of operating in a maturing, then a mature, business.  This means that they have been able to optimize away many of their traditional functions.  Such as engineering, research and development.  This would have served them well if the world had had the decency to stay constant.  Unfortunately, this is not the case, and they desperately need engineering muscle.

Telcos need a strong ability to do research and development and to have the ability to make stuff.  This is the antithesis to the core values held in high regard in many telco operations.

The reason for this is quite simple:  if you look at the sort of companies that pose the greatest threat to telcos, they are all technology companies. Facebook is a technology company.  Google is a technology company.  Amazon is a technology company.  Companies that have recognized that it is neither fast nor cost-efficient to be entirely dependent on vendors.  Companies that have recognized that they have to own the key enablers of their services.  Companies that have strong internal engineering cultures and where engineering talents go to create and build stuff.

In the time it takes a typical telco to decide that they want to build a particular service, the more nimble and capable among their "new" competitors will already have done it and rolled it out.  Globally. And by the time telcos are able to squeeze out a mere baseline product, the competition is already on their third or fourth iteration and the goalposts will have moved further afield.

Postulates
  • Reducing percieved (not necessarily actual) fragmentation is an obvious way to approach the communication challenges of users today.  However, this is very, very hard to do well.  It can't really be done effectively in back-end systems alone, but must primarily be addressed "at the edges".  In other words: applications and apps that aggregate -- possibly with some back-end support systems.  The problem is: the services you want to aggregate do not necessarily want you to aggregate them.  They want to own the user.
  • Telcos have a diminishing role in providing infrastructure.  In particular service infrastructure. The service infrastructure that is offered by most large Internet players is not based on off-the-shelf systems.  Telcos have optimized away their technology capabilities.  This made sense in the 90s and early 00s -- but today it puts telcos at a perilous disadvantage.  While telcos still provide considerable amounts of low-level connectivity (ie. bit-pipes), this connectivity can easily be replaced by other players.  Things are tending towards higher speeds at lower cost so good luck making lots of money there in the future.  In my opinion telcos must:
    • Cultivate a strong engineering capability
    • Accept that it will take years to (re)build engineering capability
    • Be willing to accept failures, learn, get up, and try again
    • Take ownership of key enablers
    • Understand Internet-scale engineering.
  • The game is "global or nothing".  If you are in the communication business and your product, or its underlying infrastructure, is based on the notion of a local or national market, you are wasting your time and shareholder money.  This applies to everything.  From the user-facing services themselves to everything that need to be in place under the hood to make it happen (user management, billing, payment etc) to how the organization operates. Everything.
In essence, "everyone" knows that the world the traditional telcos address is slowly disappearing.  Everyone knows that at the end of an industry maturity cycle, engineering capability becomes vital.  The business-school literature exists and there is a plethora of examples of how once dominant players either failed spectacularly or faded away.  (I'm not an MBA so it really shouldn't be my job to educate people on these things ;-)).

This piece doesn't offer that many answers, but it provides some pointers to how I perceive the current context.

No comments:

Post a Comment