Your browser is not supported. Please update it.

24 April 2014

Comcast changes reporting rules to ‘confuse’ the picture on TV subs

The number that dominated the announcement of the Comcast first quarter results was the one that said that pay TV customers had risen for the first time in about 7 years. Comcast had the temerity to say that this was the second quarter running.

Here’s the trick the company has pulled, and it genuinely reflects some part of its business, but you have to be aware which part. Where customers where previously billed collectively in an apartment complex, or in healthcare facilities and in other multiple dwelling units (MDUs) they were billed as one customer. Now, where there is a direct billing relationship, they are billed separately.

This was explained by CFO Michael Angelakis, who said that ‘beginning this quarter we are changing our methodology for counting video customers from an equivalent billing units or EBU approach to a billable customers method.’ What this means is that there remain bulk property agreements in place but where Comcast can get any extra revenue, it counts it as a new customer.

And what that has meant is that as the Xfinity X1 has rolled out to the customer base, including apartment blocks, where the content and EPG is delivered over the DOCSIS signal, this is all billable separately for the first time. It is this factor alone that has led to the average revenue per customer falling ‘ if you count an apartment block as one customer and it pays several thousand dollars a month, that drives up ARPU and if you count each individual customer once again for smaller increments of revenue, that drives down ARPU.

Another number that Comcast pushed is the one that it is shipping between 15,000 and 20,000 X1 boxes per day and says that this number has doubled over 6 months as the company has got better at selling it. We think that many of these MDUs have only had basic cable, in some cases still analog, for a long time ‘ and if you show them the modern OTT style X1 interface, they are going to go gaga for it, because it is such a big leap of technology.

We’re okay with Comcast claiming this means that it has actually increased its pay TV customers, even though it either hasn’t or it hasn’t by very much. At the results conference executives claimed that this was the second quarter of growth, but it did not announce pay TV customer growth last quarter, so this was a retrospective calculation based on ‘if’ Comcast has used this process last quarter, and secondly it also said that if you reversed the change, then the growth was just 4,000 TV customers (hard to believe) ‘ still the first quarter in 7 years that Comcast can claim TV growth.

What this implies is that Comcast is beating off the threat of Netflix, which it is not (see Netflix figures) because what is DOES hide is how many customers cancelled net of those who were already Comcast customers, but which by having a direct relationship with Comcast for the first time, appear to be new. We suspect that the overall message, that X1 is a good thing, has to be diluted with the fact that video revenue went up less this quarter than in previous quarters.

For the whole of 2012 video revenue rose some 12.6%, so around 3.2% each quarter, and for 2013 is rose 8.3%, so just under 2.1% each quarter. The video revenue rise at Comcast was 1.3% this quarter, almost half of what it has been.

The executives answering analysts tried to imply that this is both deliberate, and that it is not a trend, and that video price increases were under Comcast’s control. We cannot see how, given that this is the adding of low end MDU customers that were already Comcast customers, most of them at the bottom end of the pricing scale. Comcast reported 21.6 million video customers in the 4th quarter, and some 22.6 million this quarter. It has restated Q1 a year ago in 2013 as 22.8 million subscribers, even though it actually reported at the time 21.9 million.

So the new reporting process has added about 1 million video customers to a revenue figure that has hardly moved, so it is bound to have that kind of negative ARPU effect ‘ it’s basic math ‘ divide the same figure by a much higher number and the result is lower. Duh!

Comcast execs said that this way of reporting is more transparent, but that seems laughable given that no-one else has managed to work this all out and explain it in print, and headlines abound that pay TV customers are on the rise, when that’s clearly debatable.

During the past two years a little under 100,000 customer homes each quarter cancelled Comcast TV service, so we have no reason to expect that this will have gone down much – X1 customers may be driving up loyalty, but even at 20,000 new homes per working day an entire quarter only means that 1.2 million have been upgraded so this would not affect churn much across 22 million customers.

Transparency would have been telling us how many homes that already get Comcast TV service are now directly billable, and allowing us to calculate how many homes have actually left Comcast for TV service. Our guess is that 124,000 customers have been created during this quarter, who were in group deals at MDUs and that 100,000 homes have left the Comcast TV service, leading to a 24,000 rise in customers. But as we say, when you have to guess, it means there is less, not more, transparency.

This is especially true given that there was no number given for X1 homes in total and that for the first time since it was launch the word ‘Streampix’ the additionally priced movie service which Comcast launched as a Netflix spoiler, has not even been mentioned on the release or during discussion. We are left asking what would happen to a customer that was part of a group TV service in an MDU, who also bought Streampix, and the answer is that we don’t know ‘ it may have been counted as a new home or it may not ‘ Comcast is not saying (although we suspect it would be counted). The only thing in Comcast’s favor is that it claims that all other US companies already count numbers in this way, and of course this is really about the fact that all of this would come out as the FCC and the Justice Department digs into whether or not to allow Comcast to acquire Time Warner Cable. Effectively Comcast was undercounting homes by around 1 million homes and it needed to eliminate this, so it found a nice way to do it.

It also explains why in this year of all years that Comcast has not gone for the jugular with a massive pay TV price increase ‘ because that would also have legislated against it as the merger is considered. It will instead save up its price increases until after the merger is approved or not.

Conveniently this is likely to mean that given around 3 million homes are in MDUs out of the 22 million Comcast reaches, that as it develops more direct relationships, it will rise in customer numbers for about 3 years in total or another 2 years into the future, and then the inherent fall in its homes will overpower this artificial increase. That’s based on Comcast having around 15% of its customers living in MDUs, which we are told is the US average.

We don’t think it’s a complete lie that this is about improved products, but that’s only a small part of the picture.

Another key number from Comcast is that it has now rolled out about 8.3 million WiFi devices, this is against a previously disclosed 6 million a few quarters ago, so there is a rapid thrashing about under the surface which involves pushing out as many WiFi based devices as possible to customer homes ‘ but this, we estimate is still behind the number of WiFi devices that AT&T has, even though it has less broadband lines, and they are of less capacity than Comcast’s. In another few quarters we would anticipate that at this rate Comcast will have the most WiFi devices under its network control, of any player in the US, overtaking our estimate for 10 million at AT&T.

The company executives refused to allow anyone to get away with the ‘assumption’ that it would either go into business with Verizon or anyone else as an MVNO, nor launch WiFi first attachment services (as other cable companies have done in Europe). And would only add that 75% to 80% of mobile data consumption is happening on Wi-Fi in the home and in the office. Basically the company is saying ‘we are aware of what a wonderful asset this is, and no we have not yet said how we will use it.’

But the chances are that a WiFi First MVNO with either Verizon or Sprint (we think Sprint) is on the cards in the very near future, using those WiFi assets to the full.

Finally the broadband additions of just 383,000 were lower than in previous quarters, probably a shortfall of 50,000 or so, and the increases in pricing in cable broadband are finally starting to take their toll, so its broadband pricing may need tweaking downwards.

Comcast said it was happy with all the numbers, but right now everything that the executives says must be seen in the light of the merger with TWC, upon which many reputations rest. Our view is that it will and should be denied.