Back in 2007 I wrote a post about why I think that the 'Average Revenue per User' (ARPU) has become totally meaningless and needs to be replaced by another measure. Three years later, not much has changed on the reporting front, but I think in terms of definition, the general ARPU is even more meaningless now as back then. In fact I think the lower the ARPU reported by a network operator is today when compared to other national operators and operators in comparable countries, the better it is for the network operator.
The lower, the better!? Look at it this way: The lower the ARPU, the more subscribers have been won over by the network operator to use the network that do not spend as much as subscribers that use the network more. But even though those subscribers spend less, they are by no means less valuable to have. Also, there's a growing trend of people owning more than one SIM card for various reasons such as making cheaper calls to other networks with a second SIM card (hello Dual-SIM phones), an additional subscription for business, a 3G stick for business use, a 3G stick for private use, additional 3G enabled gadgets such as the iPad, 3G enabled netbooks, etc., etc.. All require a separate SIM and hence a separate prepaid or postpaid subscription that generates its own ARPU.
From this point of view I generate at least 5 separate ARPUs today and even more if you count the various prepaid SIMs I occasionally use for test purposes. So the PU "Per User" part of the ARPU is far from reality today. It's really time for a new measure…
Martin,
Check out Dean Bubley’s post on this subject as well: http://disruptivewireless.blogspot.com/2010/08/telcos-time-to-stop-reporting.html
Hi Tsahi,
yes, Dean and I share the dislike for the ARPU definition and reporting 🙂
Thanks,
Martin
Wasn’t ARPU originally intended to be an estimator of profitability? Marginal costs should rise slower than marginal revenues, so rising ARPUs implied rising profits. As carriers take on more pre-paid customers and as subscribers use more (low margin) data, isn’t ARPU still useful in that sense? Seems to me that ARPU can be interesting as long as it’s taken in context, e.g., along with the subscriber/SIM growth rate. Falling ARPU in conjunction with negligible growth indeed would be cause for concern.
Incidentally, if more countries would abandon calling party pays in favor of mobile party pays, then there would be no need to own multiple *domestic* SIMs. This could be a net benefit for carriers who would wind up keeping more airtime revenues in-house. (And the fuss about “interconnect” rates would go away, too.)