& cplSiteName &
Comments
Threaded  |  Newest First  |  Oldest First        ADD A COMMENT
iainmorris
iainmorris
1/9/2017 | 2:40:46 PM
Profits
So the implication is that operators are losing money in selling mobile data services (if costs are higher than revenues, you are making a loss). And I don't quite get that as the majority seem to be very profitable companies (even if their revenues and profits are in decline).

I also don't really get the economics. We are led to believe there is soaring demand for mobile data services and always being told about a capacity crunch on networks, which is spurring the need for new technologies and access to spectrum (itself a finite resource). There has been some consolidation in the market, too - even if there is regulatory resistance to M&A activity in Europe. 
Carol Wilson
Carol Wilson
1/9/2017 | 2:48:00 PM
Re: Profits
The issue is that operators haven't been able to increase the prices consumers pay on a per-bit basis so as consumption goes up but, relatively speaking, prices don't keep pace. Every major operator I know has been discussing this for a while now, it's what fuels the whole SDN-NFV push.

And it gets worse as traffic continues to ramp. Traditional ways of building networks don't allow for the massive scaling needed to handle the growing bandwidth demand. That's why operators want to adopt the same kinds of technology that FB and Google are using - commodity hardware controlled by intelligent software - but they can't do it fast enough. 

So operators are cutting back on capital expenses to preserve their profit margins, and will do so until they have a way to ramp up capacity that is less costly than today's methods. 
SVPCloud29186
SVPCloud29186
1/9/2017 | 2:51:39 PM
Re: Profits
The study referred to at Verizon was looking at the backbone costs of transport, not the access, and certainly not 4G and beyond.  The way I think about it is that now the access services (broadband, video, advertising) must subsidize the transport network.  Profits are still generated, but there is incerasing drag on them. 
Gabriel Brown
Gabriel Brown
1/10/2017 | 3:33:54 AM
Re: Profits
I'd be very cautious about this claim that the cost to transport bits is greater than revenue. It's not spelt out in the article or the blog post what is referred to. Does he equate revenue to the retail price?. As Iain says, many (most) operators are profitable. 

The only way this might make sense is SVPCloud29186's comment:

The study referred to at Verizon was looking at the backbone costs of transport, not the access, and certainly not 4G and beyond.  The way I think about it is that now the access services (broadband, video, advertising) must subsidize the transport network.  Profits are still generated, but there is incerasing drag on them.
mcande
mcande
1/9/2017 | 5:10:44 PM
usage based billing
Why can't the industry make the transition to usage based billing. If consumers pay for what they actually use, the revenues grow organically as the usage goes up.
brooks7
brooks7
1/9/2017 | 5:22:00 PM
Re: usage based billing
 

All of the plans - SDN, NFV (usage based pricing) are all about increasing the price of service.  What I think there is little discussion of is the value that customers are going to derive from that increase in pricing.  By trying to raise the value of Enterprise cost per bit per second, the carriers are going to drive the bulk of traffic into large data centers where Enterprises will simply bid out to the lowest bidder the required bandwidth.  Consumers will simply complain to the FCC who will probably slap on price caps.

So, if you want adoption of all this modern stuff...figure out how it will drive down Enterprise spend.   Because THAT is the value that people will pay for.  You won't raise the top line, but the goal is bottom line - so less expansion to make the revenue number.

Other than that, you will see ongoing consolidation plays.

And Telecom Equipment has been in a depression since 2000.  It isn't coming back to where it was in the hey day of the CLEC/dot.com days.  The money has moved into other places.

seven

 
mcande
mcande
1/9/2017 | 5:27:50 PM
Re: usage based billing
Where consumers will see the value is by eliminating the tiered pricing. They will be willing to pay for a better experience. We have been doing it for years with natural gas, water, electricity, etc.
brooks7
brooks7
1/9/2017 | 5:37:00 PM
Re: usage based billing
 

No fast lanes...Title II remember.

 

And Carol, all the SP would have to do to increase revenue is to throw away more packets in WRED.  

seven

 
Carol Wilson
Carol Wilson
1/9/2017 | 6:35:16 PM
Re: usage based billing
Okay, I'll bite - how does throwing away more packets in WRED generate more revenue?

 
brooks7
brooks7
1/10/2017 | 1:04:21 AM
Re: usage based billing
Carol,

If there are retransmitted packets, one might count them against the usage cap.... :) 

Edit:  I just realized that you might not know that there is an anti-congestion protocol that tosses away random traffic.  You could have it on and then change the parameters anytime you wanted to make more traffic happen.  Not good throughput, but just bits.

So, how about this...all Internet Service is free up to 1 TB of traffic.  That number doubles every 2 years.  After that you can charge.

seven

 
Carol Wilson
Carol Wilson
1/10/2017 | 11:37:43 AM
Re: usage based billing
For the record, I was aware - albeit vaguely - that something called weighted random early detection existed, I just hadn't thought about it much, and especially not as a revenue tool. 
brooks7
brooks7
1/10/2017 | 6:37:09 PM
Re: usage based billing
So, Carol...if the operators are claiming that they are losing money on bandwidth...what are they making all their profit on?  Verizon and AT&T both made over $3B in profit last quarter....so its coming from something....

seven

Edit and PS:  Carol the D stand for DISCARD not DETECTION.  Weighted Random Early DISCARD.  Its the kind of stuff that will make things get real...dicey....real fast with the whole usage based pricing.  For example, ever wonder how much data is used by those video ads on your phone when you surf the web?  You didn't want the video...but you are paying for it.
Gabriel Brown
Gabriel Brown
1/11/2017 | 3:31:32 AM
Re: usage based billing
If Tom is talking about IP transit, as someone mentioned down thread, the assertion makes more sense.

There's a good piece here on "The Death of Transit": http://www.potaroo.net/ispcol/2016-10/xtransit.html It's fair to say the author is not amassive fan of CenturyLink buying Level 3.

You can't read across to thew whole of service provider business from this data point (for obvious reasons).
brooks7
brooks7
1/11/2017 | 10:42:42 AM
Re: usage based billing
 

Thanks Gabriel.

The Service Provider community is awesome at throwing up the we are in trouble card.  It is also used by proponents of one thing or another to justify what they do.  I know Tom is talking about NFV and SDN, but the transformation and change thing has been around the industry for 15 years.  And nothing happened to put them out of business or transform them.

seven

 
kq4ym
kq4ym
1/19/2017 | 11:21:28 AM
Re: usage based billing
Yes, it's always a bit scary to try to predict the future, even based on what seems to be some rational arguments and data. Whether the current data is predictive or just a randomness in the long term, we'll not know for sure until the future is behind us. It will be interesting to watch who was on the winning side of the investment vs. revenue business planning.
Carol Wilson
Carol Wilson
1/9/2017 | 6:36:00 PM
Re: usage based billing
Service providers would love to charge more for a better experience, but regulation makes that unlikely. 

 
mcande
mcande
1/9/2017 | 7:09:12 PM
Re: usage based billing
I'm not saying charge more for a better experience. When you sign up for electricity, you don't ask for a 20Amp service because you only have two light bulbs and a toaster; you get a 200Amp service and you consume only what you need and are billed accordingly.

Service providers could charge a base rate for internet and only bill for the usage. If someone only emails and does light social media, why should they subsidize the heavy video user who should be paying more for their usage. Every consumer (rich and poor) should have the best connection (experience) possible.
Carol Wilson
Carol Wilson
1/9/2017 | 5:22:32 PM
Re: usage based billing
Possibly because consumers would revolt. There was a usage-based billing push for broadband about 10 years ago and regulators and consumers alike pushed back. It turned out almost no one truly understood how much bandwidth they were using. Consumers didn't like the idea of bills that fluctuated and some of them didn't trust their service providers to accurate track their usage. 

I think it would still be a hard sell today, but maybe I'm wrong. The closest thing is the usage buckets that wireless providers offer. 
James_B_Crawshaw
James_B_Crawshaw
1/10/2017 | 7:00:24 AM
Revenue and cost per bit analysis
Globally, I think most mobile data services are now sold as bundles (albeit with tiered usage caps) rather than on a per GB basis. Given that these bundles usually also include voice and SMS I don't know how operators would work out the revenue per bit. Vodafone, Verizon and others stopped reporting mobile data ARPU several years ago for this reason. Calculating consumer fixed data revenue should be simpler if it is priced separately from voice minutes and line rental.

Usage based pricing for telecom data is unpopular with consumers as this Cisco study indicates.
http://www.cisco.com/c/dam/en_us/about/ac79/docs/clmw/Usage-Based-Pricing-Strategies.pdf

Other utilities such as electricity and natural gas are priced per usage. But for these services the cost of the product being delivered (gas, electricity) is a much greater proportion of the total costs of the service (including infrastructure) than it is for telecom operators. That is to say, the marginal costs increase significantly with usage for gas and electricity providers while the marginal costs (excluding fixed costs) for telecom operators are insignificant (inter-connect, additional electricity to process packets).

Calculating the cost per bit for telecom operators might be as simple as dividing total costs by the number of bits. But is it a meaningful metric? Surely what is more relevant is understanding the incremental cost per MB (additional capex/depreciation) and comparing this with the incremental revenue per MB. I think we would end up comparing a number that is close to zero (incremental cost per MB) with zero (incremental revenue).

Anecdotally, Swisscom and Vodafone do not appear concerned about rising data volumes as this article points out:
http://5gwnews.com/index.php/90-r/574-100-wireless-demand-increase-easily-met-swisscom

"the move from 3G to 4G LTE reduced costs [per GB] from 50%-90%. We're seeing a similar fall as telcos go from LTE to LTE Advanced".
brooks7
brooks7
1/10/2017 | 9:23:24 AM
Re: Revenue and cost per bit analysis
Well, yeah if costs exceeds revenue then why is everybody making money?

seven

 
Carol Wilson
Carol Wilson
1/10/2017 | 10:24:34 AM
Re: Revenue and cost per bit analysis
Actually the article does spell out where the information comes from - Nolle's direct survey of 47 global operators. This is what they are saying, and have been saying for some time now. They are talking about their overall costs, and it isn't broken down by access costs, transport costs or wireless versus wireline networks. They are looking at the macro trends of how they continue to scale their networks to meet demand for bandwidth, and seeing how the current method of building networks doesn't scale to meet future demand. 

 
mendyk
mendyk
1/10/2017 | 11:00:00 AM
Re: Revenue and cost per bit analysis
There's probably a bit of "woe is us" going on here, and a bit of arithmetic sleight of hand. Even so, network operators have seen this coming for years, and yet they haven't exactly rushed to change their MO to adapt to changing times.
Carol Wilson
Carol Wilson
1/10/2017 | 11:14:31 AM
Re: Revenue and cost per bit analysis
I think it depends on your definition of "rushed." Companies such as AT&T and DT have certainly been moving much faster than they did historically to virtualize their networks precisely because they did see this coming. 

Tom Nolle also shared his candid opinions on that process, which - spoiler alert - are somewhat critical. But I'm getting input from other industry sources before sharing those here. 
mendyk
mendyk
1/10/2017 | 11:25:22 AM
Re: Revenue and cost per bit analysis
Agreed -- some CSPs have a stronger sense of urgency. And pain and suffering are part of just about any kind of transformation. It's good to hear that our collective empathy for network operators may be running dry.


Featured Video
Upcoming Live Events
October 22, 2019, Los Angeles, CA
November 5, 2019, London, England
November 7, 2019, London, UK
November 14, 2019, Maritim Hotel, Berlin
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
March 16-18, 2020, Embassy Suites, Denver, Colorado
May 18-20, 2020, Irving Convention Center, Dallas, TX
All Upcoming Live Events