Putting the house in order

There are huge inconsistencies in the way mobile operators report key performance indicators (KPIs). Every quarter, when the time comes to open the covers of operators’ well-polished earnings books, we – at GSMA Intelligence Towers – are always expecting a few disruptive changes in reporting style.

It's no secret that mobile operators do not use standard definitions for every KPI, but we have noticed that when a reporting style changes it’s often either to inflate performance or to hide underperforming activities.

Take France Telecom’s Orange as an example. Until recently, the operator group reported "personal broadband customers," which used to club together EDGE, 3G and HSPA devices. But in its Q1 2009 results, Orange started reporting "3G customers" in France, Spain, the UK and Poland. A naive analyst would therefore think that they have streamlined and backdated their previous reporting to exclude EDGE and only include WCDMA devices, but said analyst would be mistaken. In fact, the new "3G customers" figures are even more misleading and still include EDGE – in markets such as Poland or Romania, we estimated that EDGE represents a very large share of the reported total "3G customers" figures. Therefore, every quarter, we have to eliminate EDGE figures; as a standalone technology nomenclature, "3G customers" should in this case simply include WCDMA and WCDMA HSPA.

But the award for the most complex reporting style in the industry goes to Vodafone. We have recently reviewed financial and operational figures reported by the operator group that were as complex as deciphering Dan Brown’s The DaVinci Code. In 2002, Vodafone had a very particular geographical segmentation that divided Europe into three zones (northern, central and southern). Then, in 2003, central and northern Europe zones were grouped under a new northern Europe zone. Later, it started to consolidate Arcor – its fixed broadband business – within its figures for Germany. Finally, in 2008, Vodafone created new regions: Africa & Central Europe, and Asia Pacific & Middle East, and also started to break out MVNOs and visitor roaming from its service revenue figures. But the challenge really began when we tried to understand if the reported figures were representative of the group’s proportionate ownership or not (reflecting equity interest). In 2003, Vodafone reported “statutory turnover” and some proportionate figures for comparison, but from 2004 onwards it seemed as if they only reported proportionate figures. Unfortunately, it is not so simple; in its latest results, Vodafone Group is reporting proportionate revenue and customer figures for most markets, apart from India, Egypt, Vodacom and Qatar (for which it shows the total sum of operations). For instance, Vodafone Group only owns 51.58% of Vodafone Essar in India, but it reports revenue figures for the entire unit – a 100% equivalent interest - as well as customer figures at 66.98%, given it assumes the exercise of call options in this market.

In addition, this year Vodafone stopped reporting 3G devices. Our guess is that in markets such as Italy the operator has been reporting a customer loss for the last four sequential quarters, which might imply that 3G customers or devices are also declining. Its rival TIM has been following a similar downward trend and has reported a drop of 122,000 3G connections between Q2 and Q3 2009.

Russia’s VimpelCom is a great example of how mobile operators should report customer figures in highly penetrated markets. Almost two years ago, the Russian operator restated its registered cellular connections base (which includes inactive and redundant SIM cards) to report active connections. The main impact was to increase ARPU and report more realistic growth and market penetration. In contrast, Cosmote in Greece is still reporting registered connections – with large prepaid volumes and declining contract connections – which has contributed to distorted customer share and penetration close to 200%. This is the most common data clean-up policy happening around the globe, and we expect to see more operators (including Cosmote) start reporting their active customer base in 2010, especially to counter the effects of user saturation in developed countries and rapidly falling ARPU.