Analysys Mason: Why keep investing in 3G instead of focusing capex on LTE?

Johann Adjovi

LTE is gaining momentum, with more than 500 deployments implemented or announced to date worldwide in 120 countries, according to the GSMA. Significant capital expenditures are required to roll out or expand LTE coverage, so a common challenge that CTOs and CFOs face is whether they should continue investing in 3G. CEOs or CFOs are eager to understand from the CTO why spending on the 3G network should continue instead of focusing resources on LTE.

LTE has many advantages over 3G from a marketing, technical and financial point of view.

  • Marketing: LTE enables the building of a better marketing proposition for mobile broadband, because the technology offers significantly higher headline speeds than 3G/HSPA technology.
  • Technical and financial: LTE has a higher spectral efficiency than HSPA and provides higher capacity per invested dollar. It therefore alleviates the concerns of the CTO and CFO about absorbing the increasing demand for data at lower costs. LTE also offers a full-IP suite, which allows the phasing out of legacy technologies such as E1 or circuit-switching, which may in turn reduce network opex.

Another key concern of the CFO and possibly the CEO about 3G investment is the assumption that any 3G investment would be short-lived if LTE adoption is rapid. However, although the CTO may also want more investment in LTE, the CTO is aware that 3G traffic is still increasing, which justifies careful consideration of 3G investments.

We believe operators should think carefully about a number of factors regarding balancing resources between 3G and LTE.

  • Handsets/devices: In developed countries, where LTE is rolled out as a mass-market technology, most devices already support 3G (in these countries there is also an incentive to reduce the number of 2G-only handsets for spectrum refarming reasons--as discussed in one of our previous articles). However, only a minority of all handsets support LTE. The LTE device ecosystem may also vary greatly depending on the LTE band(s) deployed in the country.  However, there may be a case for accelerating device migration to LTE by excluding 3G-only devices from operator device lines, particularly in postpaid-dominated markets. For example, 80 per cent of devices sold in Singapore were LTE-capable as of mid-2013.
  • Marketing proposition and customer migration: Significantly reducing investment in 3G will lead in the short term to weaker experience for 3G users, who still currently represent the bulk of the market. It is important that the 3G user experience remains high quality, so that when the CMO focuses on how to migrate the customer base from 3G to LTE and generating some (possibly limited) extra ARPU in the process, customers feel confident about the data proposition of the operator. Negative publicity on the 3G network may lead to slower user migration and possibly higher overall churn and market share loss.
  • Spectrum holdings: Operators generally have 2x15 MHz or 2x20 MHz in the 2.1G Hz band for 3G (because the normalised 2.1 GHz band is quite wide at 2x60 MHz). They may have also rolled out UMTS900, typically with 2x10 MHz. This amount of spectrum deployed for 3G is generally the main share of spectrum, which is in line with the recent importance of 3G in offered services. Therefore, the sizeable 3G capacity should continue to be used, and may even need to be further expanded because of current exponential growth of 3G data traffic.
  • Traffic offloading between networks: Even when most data traffic is generated by LTE-capable devices, it is possible to use 3G to optimise the user experience by actively offloading LTE traffic over the 3G network so that the 3G network is not left empty. Some may think this results in a degraded user experience for LTE users. In practice, it actually contributes to an increasingly effective experience for all users, by avoiding overload in hotspots and a better load management of both networks.
  • Urban/rural coverage: If a mobile operator only has high-frequency bands for LTE (1800 MHz or 2.6 GHz, for example), achieving rural LTE coverage will be very expensive, which would justify rolling out LTE faster in urban areas with high data traffic density while continuing 3G roll-out in more rural areas.
  • Voice over LTE (VoLTE): So far, no voice is possible on LTE, and LTE customers need to rely on the so-called "circuit-switched fallback" to 3G'for placing and receiving voice calls. Operators need to consider moving to VoLTE when the time is right.

The underlying issue that sustained 3G investment raises is that of technology waves; as one technology is replaced by the next, operators must decide on the right rate at which to shift investment from 3G--the existing "star" product, which becomes a "cash-cow"--to 4G, the forthcoming "star" product, in order to maximise overall revenue generation.

When deciding to move to LTE, the CMO, CTO and CFO need to build and share an agreed and clear view on how to manage device evolution, customer migration, spectrum refarming, traffic offloading, rural coverage and the timing for VoLTE. This clear multi-faceted vision will naturally result in defining the expected evolving breakdown of investment between 3G and LTE, which will in turn maximise the effectiveness and durability of network investments.

Johann Adjovi joined Analysys Mason in 2008, and has since worked on various projects related to broadband, cable networks and fibre (Next Generation Access) in Europe, the Caribbean, and the Middle East and North Africa region. Johann has also acquired regulatory expertise by assisting several regulators and operators in Europe on topics such as regulatory lobbying, core regulation (market reviews, etc.) and cost modelling.