Thursday, 18 July 2013

NBN: Which is more profitable: Fibre or DSL/Copper?

Reading vendor materials, Telco reports on xDSL, like BT's Openreach, and Ofcom's more consumer-focussed and "non-spin" research, a theme emerged, but somewhat obscured as the critical factor, costs (capital and operational), aren't revealed by anybody.

Telcos assume they can maximise revenues by incremental investment in existing assets and avoiding for as long as possible upgrading their Customer Access Network, CAN, to Optical Fibre. Vendors are only too happy to provide them with expensive systems that will take 10-20 years for them to recover investment over. If Telco decide to upgrade those systems early, the Vendors are there eagerly waiting to sell them new Fibre gear. It's Win-Win for the Vendors, but not really for anyone else.

This is an untested, unwarranted and probably faulty assumption, one that the Turnbull Plan embraces wholeheartedly.

The central issues about the economics of Fibre Optic CANs are:
  • Is there limited long-term rate of demand growth?
    • in terms of Access Rate and Volume Upload/Download.
  • long-term consumer sensitivity to total cost: are they will to keep paying more in real-terms?
    • Can higher ARPU's (Average Revenue Per User) be continually coaxed from consumers?
  • market demand elasticity.
    • Will dropping the price 10% be compensated for by an increase in total revenues?
We know from the growth in paid-minutes of International Telephone calls from 1965-2000 that the answer to all three of these questions is an emphatic Yes!, but only when the underlying technology costs are driven down by Moore's Law and the CAN is not a rate-limiting factor.

Optical Fibre Local Loops fulfil exactly those conditions. While Digital systems shoehorned on top of poor-quality Copper loops, xDSL & FTTN, do not.

There's a goldmine awaiting the first to understand and explore this advantage.

Just as in 2006, any 5-year forecast of smartphone sales and traffic would've been out by many orders of magnitude. The Big Players then were Windows Phone, Nokia and Blackberry - all now bit players. The iPhone, followed by Android, make the point that "1,000-baggers", in finance slang, are still possible in the Tech-world.

The difference in available production Fibre transceivers and speeds offered to customers is massive. Network Operators have near infinite growth potential in Optical Fibre CANs, while Copper CANs are at the end of their product lifecycle.

Whilst CPU's for PC & servers hit the heat-wall in 2003, they continue to drive down Moore's Law, but the end is in sight. Speeds of Optical transceivers keep improving with many more doublings theoretically possible.

Telco's, the sole owners of Copper CANs, and not their frequent competitors, Cable TV companies, operate on a few economic principles, whether they're currently true or not:
  • Capacity is a scare, expensive resource. [Yes it was, but only from 1925-1950 before backbone systems migrated to transmission systems, away from single wire-per-circuit.]
    • Fibre speeds are driven directly by Moore's Law - doubling performance every 1-2 years.
    • The speed of fibre links is controlled by the semiconductor transceivers at either end: the parts you can easily, cheaply and quickly upgrade. 
  • Telcos need to maximise return on existing assets, especially the CAN, over which they have a monopoly. Also known in Marketing as "Milking the Cash Cow".
    • Ask Microsoft and Apple over the decade+, since 2000, how that approach plays out.
    • Microsoft's share price has stagnated and revenues are now flat or in-decline as it has protected it's existing products at all costs.
      • Microsoft owned the smartphone market, with Win CE/Windows Phone, for a decade and then the iPhone showed consumers "how it's done" and got the price/performance/functionality right. Stagnant products & their firms fail.
    • Apple, as always, put the consumer at the centre of its designs, embraced innovation and was happy to cannibalise its own markets, not protect at all costs.
    • We know from Eastman Kodak, the Apple of its day, that this business model is extremely sound and long-lived. They got to a century before being wiped out by bad management.
  • Telcos use "Traditional" pricing models, "What the Market will bear", versus "Cost Plus" used in every other commodity consumer product. This is why ISPs, not Telcos, drove Internet adoption.
    • Telecommunications network costs are generally dominated by Fixed, not Variable, Costs.
      • The sunk capital costs, especially in civil works, is massive in total, but per-customer, surprisingly modest.
      • Meeting costs of Capital and Depreciation, while maximising Revenues, is the central Pricing challenge of all Network Operators (Telco, Mobile, TV, Cable, ...)
      • It's very easy to saturate your network by under-pricing your offerings.
      • It's then very expensive to restore service to acceptable levels. Ask Vodaphone about its "Vodafail" experiment.
    • Additionally, many services are carried over the CAN and the supporting (and hidden) backbone networks and switches, creating problems accounting for Direct and Indirect production costs (Per-Unit attributable costs versus Overheads).
      • These two levels of input/output conflation of accounting costs makes it very hard for Network Operators to figure the actual input costs of services.
      • Without seriously testing the price elasticity of the market, no Telco really knows if they've maximised revenue or not.
      • Even if a Telco knows it can temporarily increase revenues by dropping prices, it may be prudent not to. If lower prices stimulate demand growth to levels that can't be met with ordinary construction timetables, they are both unsustainable and unwise.
      • The ideal Telco pricing strategy is for each new technology iteration to become available just as the old system reaches capacity and needs updating. It's a fine-line, avoiding over-capitalising while maintaining adequate service levels.
  • The outcome of all this, relying on an infrastructure monopoly, not innovation and good marketing, is that Telcos, just like Microsoft, are addicted to unbelievably high, and unsustainable, Gross Profit Margins (a.k.a. EBITDA)
    • The end-game for Telcos that demand 50-90% Gross Margins is irrelevance and market failure. Most industries get by fine on 30% Gross Margin.
    • Think back to 1991/2 and IBM's record losses. It happens, quickly, and happens to "unassailable" dominant market players, without warning.
    • Or recall 2007/8 and the GFC. With the collapse of the US "sub-prime mortgage" market, everybody saw it unfolding, but who acted to protect themselves or take advantage of the coming "market correction"? Same deal here... Just how it unravels for them isn't yet clear.
The problem for Telcos is their infrastructure monopoly is increasingly coming under threat from surprising sources. Customers, especially the unserviced most-profitable high-end, are very inventive at finding market substitutes that relegate Traditional Telco lock-in strategies to irrelevance.

Look at the various 1Gbps active ethernet networks being rolled out in dense high-rise areas around the world. Including Google and its various Fibre Cities projects. Historically, big organisations worked around Telstra's outrageous long-distance call rates with ease. The biggest & best example was the CSIRO creating the largest VoIP network in the world, over AARNet. They were, I'm told, the Beta site for CISCO, who became global leaders in the large-scale VoIP on the back of the CISRO network.

While the Political Parties vacillate trying to appease the Telcos, the unserved, highly-profitable and very accessible Apartment market, a.k.a. MDU (Multiple Dwelling Unit), sits waiting for some cashed-up entrepreneur to take it from them, Forever.

That's something that the Coalition has overlooked: there is not only a substantial First Mover Advantage, but like Amazon and Google, an on-going price-advantage to the dominant player. Once established, the market-leader sweeps up everyone. It won't be Traditional Telcos, like Telstra or Optus, doing this.

In 5 years, Apartment complexes won't be served by VDSL and Fibre to the Basement. If the Coalition has insanely bet on xDSL/FTTN, they'll be slammed by an Optical Fibre service provider who will effortlessly take all their most profitable customers.

That's Bad Business, writ large, from the Party that prides themselves on Knowing Business and "approaching things in a very business-like manner".

They failed in providing broadband while the POTS failed around them from 200-2005, then they failed to act when Trujillo banged the table & warned them Telstra was in real trouble, then for the 2007, 2010 and 2013 elections, they've sought to under-sell one of our most important and capable Business Productivity and Social Equity tools: Broadband.

Optical Fibre is just good business, not an ideology to be protected or rejected.


  1. Steve, reading this reminded me of that prophetic genius Marshall McLuhan. The reason this piece has such resonance is that it actually speaks to the corporates. It's beyond geekspeak. It has guru aura.

    You could go banging on boardroom doors with this. You should really consider setting up as a corporate consultant in this field.

    Nev had the good sense to put this piece on our website. Please, do us all a favour and write a major and extensive analysis on the NBN.

    1. Thanks for the vote of confidence.
      I wouldn't know how to go about getting into Boardrooms.
      Says more about me preferring to eschew the limelight than them.


Note: only a member of this blog may post a comment.