Sunday, January 05, 2014

Thailand's long evolution to Long Term Evolution - the 4G Saga

While much of the world, including some countries that Thailand thinks of as less technologically advanced, are reaping the benefits of 4G; Thailand has yet to get beyond the dabbling stage and jump in with both feet.

It seemed to take an eternity to get on to 3G and at least one operator has made the jump to 4G trials.  Now the local regulator has finally (re)set the date for the long awaited auction of the spectrum and details are set to be announced in February.

Normally I would be a strong supporter of a free market and all that a regulated market should bring in the way of competition, and therefore when market forces kick in what that then means for consumers in terms of value for money.  However, in this case I am wondering if there is a different answer to the problem.

Having watched the mast alliances in Europe and regulator interventions in New Zealand and Australia I think that the antiquated Build, Operate, Transfer (BOT) agreements that Thai mobile operators are bound to should be scrapped.  There are two State Owned Enterprises in Thailand dealing with telecommunications, TOT and CAT.  CAT already operates a CDMA network and resells access through True Move.  When the BOT terms expire CAT is set to be the main recipient of the infrastructure, spectrum, and associated operations.  My suggestion is that NBTC abandons its normal approach and instead of auctioning off the spectrum it issues the same said spectrum to CAT and current operators become retailers only.

Before I elaborate let me lay out some of the characteristics of the Thailand Mobile market to help build the case for the argument:

  • Low loyalty and high switch rates - whilst the market is not that immature most of the competition is driven by cycles of low call prices to attract customers.  As a result your average Thai consumer is about as loyal as the next 55 Satang per Minute call plan.
  • High penetration through prepaid - the mobile phone penetration rates are extremely high in Thailand with a significant number of people having more than one phone.  This is still predominantly a prepaid market which does nothing to attract and retain customers.
  • No lock in - unlike many other markets Thailand has not gone for the SIM locked model.  This means consumers get the best of handset choice and regularly roll over to a new device, which is great for handset manufacturers and retailers, but does nothing to help operators keep consumers locked in for any decent duration.  
Just to be clear I am not a fan of the practices used by European and US operators with branded and SIM locked handsets but it does create another dynamic in the local market that builds a constraint onto the current approach of spectrum allocation.

Here's a test for you:  have a look at the window of your office or condo in Bangkok and ask yourself how many different masts you see.  Now do the same inside and try and count how many femtocells you have in the building to patch the coverage holes.  The simple fact is that the best cell sites are at a premium and so we see a lot of clutter so each operator can claim to have good coverage.  Each cell site increases operation and maintenance costs, electricity/diesel consumption, and rental fees.  All of which get passed on the you the consumer.  In Europe many operators have gone into partnership and Joint Ventures with their fellow operators to share cell sites and infrastructure.  This means a single mast will hang the transceivers for more than one operator.  Vodafone and O2 in the UK ended up decommissioning 2000 mast sites.  Not only did this bring bottom line savings to each operator but it also made a positive contribution to the environment and the people living near those sites.

Long gone are the days when operators should be competing on network coverage.  It's not enough to simply be good at operating and maintaining the mobile network, modern consumers want true innovation and differentiated service, not simply the highest number of bars for signal strength (although it goes without saying that there should be at least one bar as the mobile phone is the 4th essential service after water, electricity, and television).  By moving to a single network infrastructure company, akin to Chorus in New Zealand, access will be ubiquitous and equivalent.

An additional element of 4G is the finite nature of spectrum, at the end of the day nobody is making anymore of it.  Therefore, the future mode of operation has to be a truly efficient network and this will mean maximising specturm usage.  In this article from Bloomberg you can see the outline plans for spectrum release.  Existing and unused 25 MHz blocks will be released to create 4G spectrum.  This raises concerns on the efficacy of that idea.  As we know the constraint of an operator is the slice of the spectrum they have access to and therefore how many subscribers they can carry in that spectrum slice.  If the blocks are not contiguous then the operator can suffer edge conditions that does not allow them o make optimal use of that spectrum before bleeding out.

When you consider this within the context of the market characteristics an operator can be left with not enough bandwidth for its subscriber traffic, or conversely in the case of CAT and DTAC too much that they've paid for but aren't using.  This is not an efficient and optimal way to run a network.

Now if we move to a single infrastructure model with today's operators becoming retailers under an MVNO model where they buy minutes of usage and run as a mobile virtual network operator, that problem goes away.  CAT as the logical operator can mange the spectrum and each retailer can buy the right level of access it needs to fulfill current demand and enough for growth.  This will not lead to any waste of such a finite and precious resource as radio spectrum.

It will also do away with so much of the frustration in the market.  Too much time is spent with operators and the regulator suing and counter-suing over market "irregularities" such as interconnect charges not being paid, or too much being charged.  With a single network there will be no interconnect within the 4G space and this will free up the regulator and the operators to focus outwards on market conditions and customer innovation.  A well run and regulated market set up this way should also be good for consumers as there will only one bucket of sunk cost which would make access charges equivalent and easier to monitor and control.

Then competition will be about more important factors such as price and value, service quality and true market innovation to attract and retain customers.