The Base Case

1.1 Introduction

Currently no framework exists in the UK allowing mobile users changing networks to retain their mobile telephone numbers. This paper discusses the development of the UK mobile telecommunication market if this situation continued over the next 10 years.

The UK mobile industry has four competing network operators who have around 7 million mobile subscribers. This number will increase rapidly over the next few years, to 19.5 million subscribers by 2007.

1.2 Expected growth of mobile consumers

Figure 1.1 indicates the growth of mobile users over the past few years, and the expected growth in the UK market up to 2007.

Figure 1.1 Growth in UK mobile subscribers

This estimate has been taken from current industry forecasts to 2000 and adjusted according to Ovum estimates. The forecast is based upon expected income and expenditure levels by the potential mobile subscriber market.

The forecast shows a penetration rate of about 30% into the UK population by 2007. The forecast used in this study is conservative according to Ovum's research. Ovum's independent forecasts suggests the penetration of mobile subscribers will be higher with 20 million subscribers by 2000. This rapid growth in penetration has been experienced in the Swedish and Finnish markets, where mobile phone penetration is already above 30% and expected to reach 50% by 2000. By using a conservative forecast, we will underestimate the number of subscribers switching.

The underlying trend for mobile adoption can be classified according to four main types of use. These are:

• business; classified according to the size of the workforce as either SME's or large businesses

• security; whose users subscribe to mobile phones in case they are required for emergencies

• personal mobility; whose users have mobile phones for the convenience it offers, but also subscribe to fixed line phones

• personal residential; these residential subscribers substitute mobile phones for fixed line phones

The last three categories are personal users. Figure 1.2 indicates how these market segments have developed since the mid 1980's.

Figure 1.2 Mobile Applications

In the business market we expect growth to continue through the introduction of value added services. The development of handsets which allow roaming within and outside offices will also support the increased penetration of mobile phones into the business sector. However we believe the business market has already reached its peak rate of growth.

The main growth in the mobile market is expected to come from the last two categories during the 1990's. The personal user represents a mass market for mobile communications whose characteristics are different to the niche market mobile operators catered to in the early introduction of mobile telephony. The drivers for the new market are :

• a requirement for continuous contactability and control over the communications services by the mobile subscriber. This requires greater customisation of the service towards the mobile user

• a perceived need for instant calling, driven by the awareness of crime and other reasons

• a requirement for a 'one number per lifestyle' concept. This is similar to the 'one number per person' concept, but differentiates between the mobile number a users may have for work and social activities. The value attached to mobile numbers vary by life style, hence the adoption and churn characteristics of the two sectors will be different.

The technological inhibitors to the growth of the mobile subscribers in our forecast are:

• restricted coverage in rural areas and buildings

• a perception that the reliability and call quality is low.

These inhibitors will reduce quickly over the next few years.

The high price of current mobile calls is also a barrier to their adoption. The high call charges are partly used to recover handset costs, however this distorts calling patterns. The level of calls from the fixed network to mobile subscribers is low due to the high interconnection rates. As these charges reduce in the future, we can expect less distortion in calling patterns. This will partly drive the adoption of mobile phones amongst residential users.

We expect the high call charges to create less distortion in the business market, since call prices are less important issues. For this reason, as well as for the previously discussed differences, we differentiate between the business and personal market in our study.

1.3 The business and personal markets

We estimate the current level of mobile phone penetration in the business sector in Figure 1.3. It indicates that around 50% of mobile subscribers are businesses. This level of business penetration is in line with industry estimates and our experience of business usage in Sweden.

Figure 1.3 Mobile phones in businesses

(1) From the Government Statistical Service
(2) Ovum estimates

The estimates of phone penetration into business units of different sizes will be clarified in our user survey. Small business are classified as firms with between 1 to 49 employees, and large businesses have 50 or over employees. We estimate how mobile phone penetration will increase in both business size categories until 2007, and compare it to the growth in personal users in Figure 1.4.

Figure 1.4 Growth in mobile users by type (in millions)

We see that:

• the SME sector constitutes a significant proportion of the business market, over 85%

• personnel users will constitute the main proportion of mobile subscribers by 2007

1.4 Switching in the mobile industry

The mobile industry experiences a high churn rate. Over the past few years this has typically been between 18% to 24% of subscribers according to both Cellnet and Oftel figures. A certain percentage of mobile subscribers who leave one mobile network will join a competing network. Cellnet1 and One-2-One estimate around 30% of leavers will join new networks. Figure 1.5 describes the movement of mobile subscribers over the past few years, and estimates the number of mobile subscribers changing networks.

• 1 Cellnet's response to Oftel's March 1996 consultation paper on mobile number portability
• 2 'Market Information Update', Oftel
• 
  3 Mercury Personal Communicatin's response to Oftel's March 1996 consultation on mobile number portability

Figure 1.5 Movement of mobile subscribers (in millions)

(1) A churn rate of 20% was assumed for 1996

Figure 1.5 shows that around 1.3 mobile customers have switched between networks between 1992 and 1996. Hence 18% of current mobile customers have switched networks. This is a conservative estimate. The actual percentage will be higher if the number of mobile customers who have switched between networks prior to 1992 are taken into account.

This estimate of the churn level accords with an analysis of Orange subscribers. It has been estimated that approximately 50% of current Orange subscribers have used competing network providers prior to joining the Orange network. Figure 1.6 compares the number of Orange subscribers who have used other networks against the number of subscribers changing operators estimated in Figure 1.5.

Figure 1.6 Origin of Orange subscribers (in millions)

(1) number of subscribers from other networks based on 1995 trend
(2) Orange network launched in 1994
(3) assume 50% of Orange subscribers are ex-subscribers of other networks
(4) From Figure 1.5

Figure 1.7 shows that increasing levels of users leaving networks are joining Orange. It also indicates 50% of current subscribers who are switching networks are joining Orange - a relatively high proportion. This could be because the current level of churn or level of mobile subscribers who rejoin networks could be underestimated.

The experience of Orange indicates that users are willing to change network providers based on their user requirements and the tariffs charged by network operators. This has helped Orange achieve its high initial growth.

1.5 Switching Forecast

We estimate the number of users switching networks based on current industry churn rates. The churn rate of personal users is higher than business users, 25% and 10% respectively according to Cellnet. Due to the different characteristics of the business and personal user, we produce different switching forecasts for both these sectors.

The potential movement of mobile customers in the UK market seeking alternative service packages over the next 10 years is very high. It is possible that:

• convergence in retail tariffs between the different network operators. This will decrease the incentive to change

• greater commitment of personal users to the 'one number per lifestyle' concept may decrease switching levels ( without number portability)

We account for both of these effects by decreasing the overall churn rate in the personal and business sector. We base our forecast on conservative assumptions. These are:

• that there is a 25% churn rate in the personal market, which reduces to 15% after 2000

• there is an 8% churn rate in the business market, which reduces to 5% after 2000

• 15% of personal subscribers who leave one networks will get reconnected to another network, compared to 95% of business subscribers

• a proportion of users who switch will be driven by the closure of analogue networks. Currently there are 3.4 million analogue users in the UK with the number of analogue subscribers declining. These users will have to subscribe to digital networks by 2005. If we assume only 10% of analogue users switch networks during the conversion process (most will opt to remain with the same network in order to retain their number), this amounts to only 0.34 million users switching during the conversion process. This is a minor value compared to the churn rate due to other reasons, so does not have to be considered as a separate issue.

Figure 1.7 describes the switching forecast based on these assumptions.

Figure 1.7 Movement of mobile subscribers to 2007 (in millions)

The movement of mobile subscribers from one network to another will increase to around 0.5 million annually. The total number of subscribers who switch networks between 1996 and 2007 will be around 6.5 million.

Hence a total of 40% of mobile subscribers will have switched operators between 1992 and 2007. We suspect that the total number of subscribers who switch will in fact be lower due to some subscribers switching more than once during this period. None the less 40% of mobile telephone numbers used during this period will be changed as subscribers switch networks.

• 4 This includes the 1.3 million subscribers who have already switched between 1992 and 1996

1.6 Forecasted revenue per mobile subscriber

The average revenue per mobile subscriber will decrease. This is due to the lower average spend by personal mobile users compared to business subscribers. Business users of a mobile phone may also use it for personal calls, making it difficult to segment the market clearly. However we estimate the revenue from business subscribers and personal subscribers over the next 10 years separately in Figure 1.8.

In estimating the revenue from the business market, we have assumed that:

• revenues from mobile subscribers during the early 1990's are indicative of business usage

• the annual revenue of from SME mobile subscribers and large business subscribers is approximately the same.

Figure 1.8 Business versus personal revenues

(1) Ovum estimate

Figure 1.8 indicates that:

• mobile phone revenues amongst business subscribes decreases slightly. This is due to higher penetration of mobile phones into business users with lower levels of usage

• mobile phone revenues from personal users decrease as it becomes a personal / residential communication tool.

Click here to return to contents

2 Technological Trends in the Mobile market

2.1 Spectrum allocation

The rapidly increasing number of mobile users will absorb the available spectrum. This may result in network congestion, causing mobile subscribers to switch to operators with greater capacity. Whether congestion actually occurs depends upon:

• whether operators split cells to cover smaller areas with dedicated mobile bases. Splitting a cell will effectively double the number of mobile users who can make calls within the same area.

• the emergence of new compression technologies which may boost the capacity of mobile networks

• how quickly the spectrum used for analogue handsets can be reallocated to digital handsets. This allows a more effective use of spectrum. Mobile operators have agreed to close their analogue networks by 2005, however most expect to close their analogue networks by 2000. This will increase the available spectrum.

It seems unlikely that congestion will occur for these reasons. If it does occur, it will cause subscribers to switch between networks. We take a conservative view and assume no congestion will occur.

2.2 UMTS

UMTS is a new cellular technology which an expected introduction date after 2002. UMTS will provide GSM interworking and broadband capabilities servicing the high value mobile market. At launch, UMTS will be more expensive than an established GSM network, but prices will decline by the year 2007.

Initially corporate customers will be the main subscribers to switch. We estimate large business will switch at a conservative 5% an annum to 2007.

2.3 Dual mode handsets

Dual mode handsets which can operate on GSM and DCS 1800 frequencies will be introduced into the UK market during 1997. It will give mobile subscribers the opportunity to switch between networks depending on their usage requirements without changing handsets. This will be particularly useful for PCS subscribers requiring international coverage. Since the handset does not have to be replaced, the overall cost of switching will reduce. This may either:

• increase the rate of switching between networks as subscribers no longer pay a premium subscription price to subsidise their handset when joining networks

• decrease the rate of switching between networks since all network providers will have international roaming agreements which provide equal coverage.

We do not believe dual mode handsets will affect switching for international and national coverage issues due to two reasons:

• users select their network provider based on their coverage requirements, and have little incentive to switch afterwards

• GSM and DCS 1800 will have comparable coverage levels towards the end of 1998

There may be switching from DCS 1800 to GSM operators for other reasons though. Currently Cellnet and Vodaphone, who both operate GSM networks, have been allocated some DCS 1800 bandwidth for new services. This is to support the introduction of mobile handsets for use inside and outside the office, and possibly the residential sector. It is not to allow direct competition between the GSM operators and DCS 1800 operators, but could encourage migration towards Vodaphone and Cellnet due to the advantage of having a single operator providing coverage in both frequencies.

We account for this by increasing the large business switching rate by a conservative 1%. Figure 2.1 incorporates our estimates into the forecast for mobile switching.

Figure 2.1 Adjustment of movement for technological trends (in millions)

This increase the total number of subscribers switching networks by 0.28 million between 1996 and 2007. This increases the total number of mobile users who will have switched networks by 2007 to 8.1 million. Hence 42% of mobile subscribers will have switched networks by 2007.

Click here to return to contents

The Switching Model

1 Introduction

This paper discusses how the mobile industry will develop over the next 10 years if number portability (NP) is introduced at the beginning of 1999. NP is expected to increase the number of users who switch between networks since it allows mobile subscribers to retain their current telephone numbers.

Nearly all subscribers who switch due to the availability of NP will rejoin another mobile subscriber, whereas currently only 30% of churn subscribers do. These subscribers represent a segment of the mobile market who are more reliant upon mobile phones. They also represent a more attractive target market for mobile operators. It is these subscribers who will benefit most under NP.

2 Market size

The number of personal or business mobile subscribers has been calculated in the Base Case discussion paper. This is broken down by type of technology (either analogue or digital), for the three users groups; personal, SMEs and large business. We need to estimate the level of switching between analogue and digital networks for these users groups. All analogue users will have switched networks by 2005, but NP will increase the rate of switching.

By the end of 1996, around 50% of mobile phones were either GSM or DCS 1800 based. Based on this, and Figure 1, we can estimate the penetration of digital mobile phones into business.

In Figure 1 we assume that:  

• all large firms have digital phones ( either GSM or DCS 1800)

• around 60% of business phones are digital, slightly higher than the total digital level for all mobile phones

Figure 1: Penetration of digital phones into the business market 

Figure 2 indicates the breaks down of analogue and digital phones by user group. The number of mobile phones by type of user group are taken from the base case paper. In making the estimates we have assumed businesses will stop using analogue phones by the end of 2000 and personal users will stop by the end of 2005.

Figure 2: Number of users by type of technology and customer type 

• If all users buy digital handsets from 1997, only a total of 3.37 million analogue users will have to change to digital networks. A proportion of these will remain with their current network provider in order to maintain their existing number. Under NP, we can expect a greater fraction of them to switch networks as well. We use this in our switching model.

3 Market for switching

The different networks currently available to business and personal users are analogue, DCS 1800 and GSM. The introduction of UMTS around 2002, and will cause further migration between networks. The general movement of subscribers will be away from analogue towards digital networks - but this will depend strongly upon the costs involved in switching.

Surveys have indicated that mobile subscribers who do want to switch networks find the inconvenience of changing numbers a barrier - partly because of the costs it imposes upon them. The movement of mobile customers will depend upon the type of subscriber, either business or personal, and the different types of networks available to them. This may be expressed through an economic expressions, whereby the net benefits (Nb) acquired by the switching subscriber outweigh the net costs (Nc). A subscriber will only switch when the net gain (Ng) is positive.  

The variable i and j represent the different factors contributing to the overall benefit or cost equation (such as tariff and coverage factors). The Nc can be thought of as being composed of two components. The cost incurred in changing networks under NP, and an additional cost component which arises when changing networks without NP. This latter cost component covers the cost incurred by the subscriber of informing customers and friends of a number change.

We ignore the latter cost in this paper due to the availability of NP. Clearly this will result in a reduction in the cost of switching - increasing the Ng. This will lead to an overall increase in the churn rates. The Ng will depend on the cost and benefits each factor (i and k) will contribute. This section attempts to briefly discuss what these factors may be, and how they will affect the switching patterns at a general level.

It has already been indicated that business users have different usage and cost characteristics to personal users, and choose mobile packages accordingly. We estimate the impact of NP on the business and market sectors separately. We also assume that current analogue and digital mobile subscribers NP will have access to NP when it is introduced.

The introduction of NP will have an immediate impact upon the business market. Currently the churn rate of businesses is 10%, compared to 25% for personal customers. With NP, business subscribers can switch networks without the cost of informing customers or losing business in the process. The churn rate of business can rise as a result. This is a significant change to the underlying assumptions in the base case - given the number of business users in the mobile market.

The speed of movement will depend upon:

• the incentive to switch to better quality digital networks which are feature rich. The price difference between digital and analogue network will determine the number of subscribers who switch. This price difference is largely due to the handset costs, rather than the per minute cost for calling. This difference will reduce as economies of scale are achieved for digital handset production. (We assume business will switch to digital networks at a quicker rate than personal users. By 2000 we assume no business will be using analogue handsets, and personal users will stop using analogue networks by 2005. Some estimates indicate full digitisation may occur as soon as 2000, our estimates are conservative by comparison).
• the coverage available from different operators. Currently the GSM operators have the greatest coverage, but DCS 1800 networks will have comparable national coverage by 1998, as well as significant European coverage. This will decrease the incentive to switch between networks for coverage related reasons after 2000. But users on networks between 1997 to 2000 may switch due to coverage related reasons depending on user requirements. We do not adjust the rate of churn figures to include switching for this reason due to the difficulty of estimating such figures. We simply assume all subscribers choose networks based on an expectation of their usage, and do not need to switch at a later point.  

These above factors also apply to the personal market, although it is not expected that the introduction of NP will heighten the importance of these factors as drivers for personal users to switch networks. Other factors which may increase the overall switching rate in both the business and personal market are:  

• the tariffing packages between the different operators. We assume the competition in tariffing packages over the next 10 years will be as competitive as today, hence switching levels between the networks will either remain at the same level or increase due to NP. We assume it will remain the same.

• the introduction of dual mode handsets. These handsets will be able to inter work with DCS 1800 and GSM networks depending on the SIM card inserted. The reduction in handset prices will make dual mode handsets more accessible. Combined with NP, there will be greater switching by mobile users as they seek better tariff packages. It is possible that the introduction of dual mode handsets may decrease switching for coverage related reasons. However we have already decided to assume users do not switch due to coverage related reasons after 1998.

• we also expect the introduction of dual mode SIM cards, which will allow the separation of business and personal call through one handset. This service is currently available, but not using dual mode SIM cards. Business users with this facility who switch jobs may want to take their personal numbers but this does not affect the switching rates already estimated in the base case.  

Figure 3 indicates the incentive to switch between networks at a general level due to these issues for the different types of customers if NP was introduced.

Figure 3: Incentive of mobile subscribers to switch 

In Figure 4 we estimate how the churn rate may increase due to the net benefits acquired by the user due to NP. We assume that all mobile users who switch due to the introduction of number portability will rejoin another network.

Figure 4: Movement of mobile subscribers 

(1) These are estimates. The users surveys will give a clearer indication of the actual values
 

4 The switching forecast

The churn rate is around 25% for personal customers and 10% for business customers, although we have underestimated this figure when estimating the number of users switching networks. Figure 5 indicates the estimate for switching without NP.

Figure 5: Base case scenario 

Around 6.5 million subscribers will have switched networks during the period 1997 to 2007 without NP. This will increase under number portability due to the issues discussed in Section 2 and 3.

The impact of these is estimated for the personal and business user in Figure 6. In estimating the churn level for both the business and personal market, we have assumed that:

• all users who switch due to number portability will renew their mobile subscription

• the churn rate will decrease to 10% for business and 15% for the personal user by 2004

Figure 6: The impact of number portability 

(1) Users who change due to NP all renew mobile phone subscriptions

It shows an overall increase in the churn rate, with a higher proportion of users who leave networks, reconnecting to other networks. This is used in Figure 7 to estimate the total number of subscribers who switch to other networks.

Figure 7: Subscribers switching under number portability 

It shows that the total number of subscribers who switch networks during the next 10 years will increase to about 13.4 million. If we include the number of subscriber who have switched between 1992 and 1996, a total of 14.7 million subscribers have switched or will switch under NP. Approximately 55% more than in the base case scenario.

On an annualised basis, this results in between 0.5 and 1 million more subscribers switching each year. Most importantly, however, is that it suggests 75% of mobile users will have switched between networks by 2007. This is based on conservative estimates as to the level of churn.

5 Implications for the Industry

The number of subscribers who switch under NP is a considerable increase upon the base case but is due to a small increase in churn levels. The most significant increase in churn levels is in the business sector over the next few years as digital and dual mode handsets are adopted.

The increase in churn levels by itself is not why the number of subscribers who switch networks is so high. But because those who do switch due to NP, rejoin other networks. As stated in the introduction, it is these users who stand to gain the most from NP.

Click here to return to contents

Final switching model

1.1 Introduction

The study developed a model of mobile subscriber switching patterns based on interviews with operators and service providers. The model represents an industry consensus of current churn levels and inter-network switching within the mobile market, and incorporates the findings from the surveys of personal, SME and corporate mobile users. This model was used as the basis for the cost-benefit analysis undertaken by the study.

The model was developed along logical assumptions and incorporated several layers of parameters. This allowed greater flexibility when measuring how switching behaviour would change if Number Portability was introduced. The main assumptions behind the model described in this section, along with the main findings.

1.2 Model Parameters

The simplistic assumptions in the Base Case and Switching Model paper were refined in the final model. The model was expanded to incorporate:

• differences between personal, SME and corporate switching levels

• the model of switching was separated by mobile technology ie analogue, DCS 1800 and GSM. This meant that the model worked on a matrix of 3x3 parameters ( the three main sectors and three main technologies)

• the model captured subscribers switching out of one technology into another. These subscribers were fed back into the model, thus the model was dynamic and all subscribers switching within the model were accounted for

• the model assumes that switching will decrease over the next ten years as competition between networks increases and tariffing and price packages become comparable. Hence it is assumed the churn rate in the industry will decrease rapidly over the next 10 years, to about 12% by 2007

1.3 Switching model without NP

The central assumption driving the model is the anticipated growth of mobile subscribers over the next 10 years presented in Figure 1.1. This represents a consensus opinion on the growth of the mobile industry.

Figure 1.1: Growth of mobile subscribers in millions

This is further broken down by the rate of mobile penetration into each sector (personal, SME and corporate). For each sector, the penetration rates of phones by analogue and digital technology have been calculated separately to allow for the higher churn rates which may be caused by analogue users. These model values are presented in Figure 1.2.

Figure 1.2: Penetration of mobile phones by sector and technology (in millions)

The level of churn will vary by sector. These are estimated from the survey results. Figure 1.3 indicates the overall level of churn used in the model, and the percentage of these who will reconnect.

Figure 1.3: Churn levels

The level churn within each sector will also vary by technology, and this can be derived from the survey results. In certain instances, the preference of certain sectors for switching to new technologies was difficult to ascertain given the ambiguity of the responses (mainly DCS subscribers who seem content to remain as DCS subscribers). In these cases, the model assumes switching subscribers will be relatively indifferent to the final network (i.e. those who switch out of DCS will switch into alternative DCS networks and GSM network equally). Figure 1.4 indicates the final switching pattern between technologies if there is no number portability.

Figure 1.4: Switching patterns in millions of mobile phones