Forecast: Mobile penetration levels still distorted

Michele Mackenzie, service manager for the WirelessInternet@Ovum Advisory service reveals some surprising new Ovum forecast figures

Throughout the late 1990s mobile operators experienced phenomenal subscriber growth, adding millions of new “subscribers” every quarter. However, this was coupled with sharp decreases in Average Revenue Per User (ARPU), a key metric in assessing quality and stability of wireless operator revenues.

Operators have since found themselves in a situation where reported subscriber volumes have exceeded real population penetration. In other words, there were more connections than people. The knock on effect of this is that apparent subscriber growth figures distort “real” ARPU (in other words the APRU generated by a user as opposed to a connection).

For the operators, this scenario has made it very difficult to assess changes in customer behaviour and in particular the propensity of users to spend disposable income on mobile services. It is therefore in the interest of operators to maintain accurate data on their users.

While operators have certainly begun to look much more closely at the subscriber base that drives their ARPU figures – and particularly at inactive connections - the question is whether they have gone far enough.

What the figures show

A recent Ovum exercise in market segmentation by age yielded some surprising results (figure 1 shows our results for four developed markets). Distributing reported mobile connections by an addressable market divided by age groups revealed that penetration was still well over 100% in some age segments.

Figure 1: Penetration by age segment at September 2002

Source: Wireless Internet@Ovum

Of course, it is possible to distribute the connections amongst the different age groups using different assumptions, but most would agree that it is the youth and middle youth age groups that drive mobile adoption. Our overall conclusion is that penetration is surprisingly high in all groups in developed markets such as Italy, Germany and the Netherlands but appears reasonable in other relatively developed markets such as the US where there is not a large prepaid market.

So have we really reached the golden age of multiple connections per user?

Of course, we accept that a number of subscribers have more than one connection. Oftel conducted research into multiple SIM card use in 2001 and found that in Italy 13% of users, mainly young people, had more than one SIM. This goes some way to accounting for the high penetration rates within the age segments, but even if we take the 13% into account our results would still show penetration rates of over 100% in the youth and middle youth groups.

Such high penetration cannot be attributed to multiple connections or at least not active connections. And the inactive connections are prepaid connections which are either redundant or can only loosely be described as active (used once every 6 months?) Operators should be applauded for recent purging of inactive connections but they may need to do more.

It is also important to remember that many operators have subsidised this growth in the market, so while ARPUs have declined subscriber acquisition costs have increased. One saving grace for Italy is that, unlike many other markets, it has not subsidised the handsets for the “inactive” connections in the market.

Why are such anomalies in base subscriber counting important?

The database purge at the beginning of 2002 was a step in the right direction and recent signs that ARPU is stabilising, even increasing slightly, in some European markets is positive.

It is important to point out, though, that the Oftel research mentioned above did not find any evidence to link the ownership of multiple SIM cards to increased usage. So if the net effect of multiple device ownership is generally not to increase usage, and therefore revenues, then the operators do not benefit. Moreover they may well have increased their costs by subsidising multiple connections.

Focusing on a subscriber base made up of active users allows operators to show a “real” ARPU, which is a more accurate reflection of a user’s propensity to spend on mobile services. It also allows for better segmentation strategy, enabling operators to employ resources efficiently on those users that are willing to spend. And this is especially important when it comes to forecasting uptake of new data services and the potential revenue from those new services.

If “real” penetration of voice services is questionable, how do we determine the real ceilings for wireless data adoption and the revenues we can expect from those services?

Related Ovum Research

Wireless Middleware: Enabling Services, Driving Change – An Ovum Report

http://www.ovum.com/go/product/flyer/WAO.htm#ternet@Ovum - An Ovum Advisory Service

Business strategy and planning – Ovum Consulting