In this note we summarise the available evidence on trends in ARPU among European mobile operators. We demonstrate the increasing trend towards stable or increasing revenue per subscriber in key markets. The end to the long downward trend in voice ARPU is clearly in sight. This new stability is derived from increasingly firm call charges and slow growth in minutes of use. Local competitive conditions may disrupt this pattern in individual countries – and we demonstrate the countervailing trend in Finland – but, overall, the pattern is clear and will probably become more so in the next few months.

More important, perhaps, the current economics look acceptable both for BT's Wholesale and Openworld divisions - this note includes some detailed financial analysis. But even at the lower price levels, we remain unconvinced whether subscriber numbers will grow as rapidly as BT predicts. (BT is now saying that ADSL subscribers will be more numerous in 2005 than unmetered customers are today!)

Nokia's recent guidance suggested a modest recovery in handset sales in 2002, followed by a strong resurgence thereafter. We think the position will be different and look for unit sales of about 450m next year, with only 3-7% growth in the years 2003-2005.

In this short note, we look at trends in mobile design and features. We show that the steady decline in size and weight is now over, and manufacturers are focusing on adding new functions, such as digital cameras, and even, in one case, a thermometer.

In this issue, Toby looks at recent evidence on UK multichannel viewing, particularly in the period immediately prior to the start of the new BARB audience panel.

Mobile operators’ marketing strategies are the primary determinant of the size of the market for handsets. The level of handset subsidy dramatically affects retailer sales. In this brief note, we look at the changes in retail subsidies in the UK during the past 6 months. We show how the emphasis on the acquisition of contract customers has meant higher levels of subsidy for the phones taken by these customers. By contrast, pre-pay subsidies have fallen.

Since the research for our report on Vodafone was carried out, the UK mobile operators have made a number of changes in tariffs. We think the net impact of these changes has been to increase average call charges. This may be the first time that the UK has seen any upward trend in prices. If these price changes stick, the impact on voice ARPU is clearly positive.

Looking ahead, the costs of buying a PC and funding a connection will act as barriers to Internet expansion. Such expenditures weigh more heavily on households in less prosperous households, where the Internet have-nots are concentrated. Although the Internet-enabled mobile phone and digital TV subscription would eliminate PC ownership as a barrier to Internet access, we do not think these will be (for the foreseeable future) access platforms of more than marginal significance.

Nokia's quarterly results statement included an estimate for worldwide global handset shipments of about 390 million. Global shipments so far this year have been:

Europolitan, Vodafone's Swedish affiliate, has just announced that its expected costs to build a 3G network will be 10bn Swedish Kroner, or about 1bn Euros.

How will this happen? XP contains hooks that link it closely to Microsoft Passport and Microsoft Messenger. These two applications are the gateway into .NET. Passport provides the central storage for user details (no more tedious entry of personal data into web forms), Messenger offers a broad set of 'unified messaging' functions. One example struck as particularly powerful; XP users can automatically see which of their list of 'buddies' is online and can contact them via instant messaging. If this isn't a direct threat to AOL, we don't know what is. (These functions have been separately available for some time, but making them an integral part of the operating system will hugely expand their use).

The UK mobile operators have made much of their honesty in ceasing to record subscribers that have not made calls within six months. This will help analysts make a clearer judgment of how many people actually use mobiles, and what the correct figure is for ARPU. But pleasure at the apparent increase in openness should be tempered somewhat. Operators are starting to make active efforts to stop subscribers becoming 'inactive'. In the last few weeks all the UK operators appear to have adopted similar policies. These policies state that the operator will take back a subscriber's telephone number unless one call is made or one SMS is sent from the phone during each six-month period.

As important, the mobile operators are keeping a much tighter rein on inventories, effectively shifting stock risk to retailers such as Carphone Warehouse. Inventory levels throughout the supply chain will be lower. Retail price levels will be more robust – improving operator margins. But we expect total sales over the Christmas period to be lower than expected because of the higher prices in the retail chain.

This report provides our model for global handset sales in 2001 to 2005. We continue to forecast 375 million units shipped in 2001. The forecast for 2002 is 470 million units. Key constraints on the level of shipments in Europe:

At current pricing levels, we see pay-TV penetration struggling to exceed 60% by 2010. And the rapid continued price inflation in the pay-TV offerings of Sky, the cable companies and ITV Digital will make even this target difficult to achieve. As our recent note on household expenditure (Time and Money) indicated, the poorest 40% of the population have very little surplus cash. Increasing prices means that pay-TV is moving even further beyond the reach of this group.

We have analysed the latest numbers from the EMC world database.  The countries covered represented 68% of year-end 2000 global subs, so are incomplete but give a fair indication of progress so far this year.  In addition to updating countries to end Q2, EMC has also made retrospective adjustments to US and China numbers (both of which make the numbers more consistent with bottom-up sources we have looked at).  The countries given partially or mostly cover Western Europe, North America and Asia Pacific, and the analysis below extrapolates the missing countries and regions.

Q1: 97m

Q2: 91m