Hello Left Hand, I'd Like You To Meet Right Hand

One of the best articles I read yesterday was "FCC Chairman Mocks Industry Claims That Customers Don't Need Faster Internet" by Jon Brodkin over on the ARS Technica site.  What elated me the most was how Chairman Wheeler used the industry's own words against them to refute their assertion that the then current state definition of broadband (4Mbps down/1Mbps up) was sufficient.  He noted how the industry's lobbying arm argued against redefining the minimum upload and download speeds a service must offer in order to qualify as broadband.  He then pointed out how each of the major ISPs fighting the updated minimum levels were ALL actively marketing speeds in excess of the new limits to their customers.  The telecommunications companies are upset because existing DSL services meet the old 4Mbps down/1Mbps up; however, they no longer qualify for the "broadband" moniker based upon the updated definition.

AT&T

AT&T offers high speed internet service via Digital Subscriber Line (DSL) and U-verse.  AT&T's DSL comes in three flavors: 768kbps, 3Mbps and 6Mbps.  Their advertising page for DSL illustrates that a 4GB video file take more than 12 hours to download on the low tier, 3.1 hours to download on the middle tier and 1.3 hours to download on the "Xtreme 6.0" tier.    Their highest level of service is the ONLY level of service capable of serving up even a single HD video stream in real time.  The "AT&T U-verse High Speed Internet" product page lists service that begin with a 3Mbps service tier and continue with 6, 18 and 45 Mbps tiers. 

As of yesterday, AT&T can no longer claim their DSL service is "broadband" as it is no longer capable of meeting the new definition for minimum level of service.

Comcast

The "High Speed Internet from XFinity by Comcast" product page lists 6Mbps, 25Mbps, 50Mbps and 105Mbps service tiers.  Their description of the "Performance" tier service which offers download speeds of 25Mbps is good for streaming online video and sharing photos.  It is not until we get to the 50MBps "Blast!" tier where they start talking about streaming and downloading HD video content, gaming online and supporting households with three to five devices making use of the internet connection.

Time Warner Cable

Time Warner Cable has a nice little "WiFi*Denti*Fier" page which lets you "Dive in and Learn How You Can Enjoy the Internet Even Better".  They recommend a 15Mbps connection for a single person watching television, streaming video and using a computer online.  Add your significant other and one additional computer into the mix, and you're now better served by their 30Mbps tier.  Bumping their model up to a family of four, where each person has a computer, and you have two video streams and the "Ultimate" 100Mbps package is what you need.

Verizon

On Verizon's "Check Out the FiOS Experience" site, they describe the 25Mbps download tier as being good for normal, every day use if you live alone or perhaps share your connection with up to two people.  They describe usage such as web surfing, reading and responding to e-mail, online shopping and using social media.   They also indicate you can stream two high definition videos simultaneously.  The FiOS 50Mbps download tier is described as being the best for your family's day to day use allowing you to share photos and files quickly, enjoy online gaming and stream up to five high definition videos at the same time.  Verizon also features 75, 150, 300, and 500 Mbps plans.

Party Line Vote

What I also find telling is how the votes on re-definition of broadband followed party lines: Chairman Wheeler, Commissioners Clyburn and Rosenworcel voted in favor while Commissioners Pai and O'Reilly dissented.  Pai and O'Reilly are the two Republicans on the commission and they sided with the industry against updating the minimum requirements to qualify as "broadband" service.  The two Democratic commissioners sided with the Mr. Wheeler in support of the new definition.

O'Reilly argues that the FCC's report indicating a 25Mbps download service level is required for streaming 4K television supports a TV format that is "...not expected to be widely adopted for years to come."  As a counterpoint to Mr. O'reilly's assertion, I would like to cite a statement made back in March of last year by Barbara Kraus who is a Director of Research at Parks Associates.  In a their study entitled "4K Today: Bringing Ultra HD to Market", Ms. Kraus is credited with the following assertion:

The price curve for 4K TVs will be similar to that of HDTVs but at a slightly faster rate of decline. While 4K is priced at the high end now, it will decline to mass-market levels over the next two to three years. Like flat-panel HDTVs, 4K TV prices will continue to decline as unit sales volume increases.

A few months later in September, Marco Chiappetta describes how Vizio's launch of a its P-Series of 4K Ultra HD televisions with an entry level price point of $999 for a 50" television illustrates 4K televisions are closer to mainstream adoption.

Contrast this with Commissioner Rosenworcel asserts "[w]e invented the internet. We can do audacious things if we set big goals, and I think our new threshold, frankly, should be 100Mbps. I think anything short of that shortchanges our children, our future, and our new digital economy..." 

Summary

The cable and telephony duopoly are fighting against the FCC's recent moves to provide more oversight for the broadband internet market.  Their excuse is that it is incredibly expensive to maintain and upgrade their existing networks and FCC regulations would place an undue burden upon their shoulders.

To me, it isn't a question of the availability of funds but rather where these companies want to spend their money.   Comcast, for example, purchased AT&T's cable television business back in 2001 for $47 billion dollars.  They spent $4.8 billion to acquire MGM in 2004. In 2009, Comcast spent $30 billion dollars for a 51% stake in NBCUniversal.  Four years later, they spent an additional $16.7 billion to acquire the remaining 49%.

AT&T clearly has some extra funds available considering they agreed to Federal conditions on approval of their $48.5 billion dollar acquisition of DirecTV back in August.  A few months later, on the 7th of November, AT&T announced that they are going to acquire a Mexican wireless provider for $2.5 billion dollars.

Back in 2005, a book entitled "$200 Billion Broadband Scandal" was published.  The book outlines how during President Bill Clinton's tenure, there was an aggressive "National Infrastructure Initiative" to rewire ALL of America with fiber optic cable.  The Bell companies at the time agreed to step up to the plate and complete the work.  It is estimated that the Bell companies netted $206 billion dollars in excess profits and tax credits and in exchange, American Citizens got DSL.

The incumbent ISPs have demonstrated that they will only invest to improve their networks unless they are forced to.  Given the limited choices consumers have, they have little if any compelling reason to do so.  I applaud the FCC for taking steps to protect consumers and move to improve the current state of affairs in the broadband world!

References

Abrams, Rachael. "Comcast Hast History of Big Deals." New York Times. 13 Feb 2014. Web. 30 Jan 2015. <http://dealbook.nytimes.com/2014/02/13/comcast-has-history-of-big-deals/?_r=0>

AT&T. "AT&T to Acquire Mexico Wireless Provider Lusacel." 07 Nov 2014. Web. 30 Jan 2015. <http://about.att.com/story/att_to_acquire_mexico_wireless_provider_iusacell.html>

AT&T. "AT&T DSL Internet and Home Broadband Service." Web. 30 Jan 2015. <https://www.att.com/shop/internet/internet-service.html>

AT&T. "AT&T U-verse High Speed Internet." Web. 30 Jan 2015. <https://www.att.com/shop/internet/u-verse-internet.html>

Broken, Jon. "FCC Chairman Mocks Industry Claims That Customers Don't Need Faster Internet". ARS Technica. 29 Jan 2015. Web. 30 Jan 2015. <http://arstechnica.com/business/2015/01/fcc-chairman-mocks-industry-claims-that-customers-dont-need-faster-internet/>.

Chiappetta, Marco. "4K Ultra HD Inches Closer to Mainstream Adoption". Forbes. 29 Sep 2014. Web. 30 Jan 2015. <http://www.forbes.com/sites/marcochiappetta/2014/09/29/4k-utlra-hd-inches-closer-to-mainstream-adoption/>.

Federal Communications Commission. "FCC Finds U.S. Broadband Deployment Not Keeping Pace." 29 Jan 2015. Web. 30 Jan 2015. <http://transition.fcc.gov/Daily_Releases/Daily_Business/2015/db0129/DOC-331760A1.pdf>.

Kosman, Josh. "AT&T Agrees to Conditions With Feds in $48.5B DirecTV Purchase." New York Post. 25 Aug 2014. Web. 30 Jan 2015. <http://nypost.com/2014/08/25/att-agrees-to-conditions-with-feds-in-48-5m-directv-purchase/>

Kushnick, Bruce. "The Book of Broken Promises: $400 Billion Broadband Scandal & Free the Net".  New Networks. 01 Oct 2014. Web. 30 Jan 2015. <http://newnetworks.com/bookofbrokenpromises.htm>.

Tretbar, Alex. "Study Projects 4K Will Expand Like HD, Only Faster". Digital Trends. 28 Mar 2014. Web. 30 Jan 2015. <http://www.digitaltrends.com/home-theater/4k-will-follow-path-similar-to-hd-growth-pattern-but-faster/>.

Parks Associates. "4K Today: Bringing Ultra HD to Market". 20 Mar 2014. Web. 30 Jan 2015. <http://www.parksassociates.com/bento/uploads/file/research/ParksAssoc-4K-SpecialResearch-TOC.pdf>

Verizon. "Check Out the FiOS Experience". Web. 30 Jan 2015. <http://www.verizon.com/home/fios/#fios-experience>

The Cavalcade of Dumb

Yesterday, Apple announced their financial results from the first quarter in their 2015 fiscal year.  It was good.  They announced a record $18.04 BILLION dollars in PROFIT for the quarter.  That is the most profit for a single quarter than any OTHER COMPANY IN THE WORLD. Ever!  Prior to this, the record was $16.2 billion reported by Russia's Gazprom.  The following are some of the other highlights are:

• 74.5 million iPhones sold (best quarter ever)
• 5.5 million Macintosh computers sold
• App Store revenue up 41%
• 21.42 million iPads sold

As Tim Cook noted, Apple was selling 34,000 iPhones an hour, twenty-four hours a day for EACH HOUR in the quarter.  Their gross margins were reported to be 39.9%.  The number of iPhones sold represent a 46% increase over the volume sold in the year-ago quarter.  The number of Macintosh computers was up 15% from the year-ago quarter.

But it's not all rosy.  iPad sales continued their downward sales trend - 21.42 million iPads represent an 18% decline from volume in the year ago quarter.  Have a look at the following illustration of iPad sales numbers by quarter since the iPad's release in the third fiscal quarter of 2010.

Disturbing, isn't it? 

Well, that is what some pundits and naysayers want you to believe.  They will insist with absolute conviction that  things are dire for the folks in Cupertino.  They will insist that Apple must take immediate action to address this manufactured "problem" facing the iPad line.  Add a stylus!  Release a "pro" version!  Drop the price! 

They just don't get it.

The iPad Air has a 98% customer satisfaction rating. 82% of all U.S. e-commerce transactions coming from tablets, came from iPads. More than 70% of all table-based web traffic comes from the iPad - according to Chitika. Of the iPads that are sold in the US, 50% go to new iPad buyers. In China over 70% of the iPad purchases are from new customers. According to IBM's review of point of sale data, iOS devices accounts for twice the online traffic and four times the commerce sales relative to Android-based devices.  Apple has sold a staggering 258.61 MILLION iPads.

iPad sales are down 18% from the year ago quarter.  Given that fact, they still delivered 21.42 MILLION iPads into the hands of consumers last quarter.  iPad usage is SIX TIMES that of its nearest competitor!

The iPad is just fine.  Apple's problem, if you can call it that, is exactly as Robert Cringely observed back in May of 2014 - the iPad is just made really, really well.  Apple does not include planned obsolescence in their products.  Apple's paramount concern is releasing the best product in a category.  Because of this, people tend to use their products for longer periods of time.

Half of the people buying iPads in the United States are buying their first iPad.  Since its inception, Apple has sold 258.61 MILLION iPads and the majority of those sales are going to NEW customers.  We do not know the average replacement cycle for iPads, but I would speculate that it is a lot closer to that of the Macintosh than it is to the iPhone.

Since 2002, because of my work as a consultant, my primary computer has been a laptop. On average, I get a new laptop every four years.  When I get a new laptop, the previous machine is used by other members of our family when they don't feel like sitting behind their desktops or when we travel.  Consequently, the useful life of a Macbook for our family is between seven and eight years.  The 17" Intel Core Duo Macbook I bought in 2006 is still actively being used by my father on a daily basis, so that machine is closing in on nine years of utility.

My wife used a 450Mhz G4 as her primary machine at home for about five years when it was replaced with a G5 PowerMac.  The two 1Ghz CPUs in the G5 desktop provided more than enough computing capacity to sustain her for seven years.  There was NOTHING that machine couldn't do!  The only reason we replaced the G5 with a 27" iMac was because the Adobe Creative Suite of products stopped supporting the PowerPC architecture.

I see our family's use of iPads following a similar trend.  I pre-ordered one of the first generation iPads the moment it went on sale in 2010.  When the second generation iPad came out in March of 2011, I ordered one of those because my wife and I discovered that any time one of us wanted to use the iPad, the other person had it.   For me, it seems that the iPad replacement cycle is two years and eight months because that is how long I waited before I got the iPad Air.  However, the second generation iPad is still used in our family on a daily basis so I would submit that it still has some legs.

In my opinion, there is still tremendous upside to the iPad story.  I do not believe Apple ever intended for the iPad to fully replace personal computers.  It will for some, but not for everyone.  What the iPad will do is replace traditional computers for people that only used computers for web surfing and light e-mail use.  Tablets in general and the iPad specifically afford us the opportunity to bring computing into more scenarios where a desktop or laptop computer may have been a cumbersome or awkward fit.

Speaking from personal experience, there is not a day that goes by where I do not use my iPad, iPhone and MacBook Pro.  Each device fills a specific need and purpose.

Still.  In spite all of this, there will be those who forsee a bumpy road ahead for Apple's tablet.  No volume of logic or reason will sway them from their pulpit of doom.  Perhaps someday they will be right.  After all, even a broken clock is right twice a day.

References

Cringely, Robert . "Apple's iPad Problem." I, Cringely. 19 May 2014. Web. 28 Jan 2015. <http://www.cringely.com/2014/05/19/apples-ipad-problem/>

Kumparak, Greg. "Apple Just Had the Most Profitable Quarter of Any Company Ever". Tech Crunch. 27 Jan 2015. Web. 28 Jan 2015. <http://techcrunch.com/2015/01/27/apple-just-had-the-biggest-quarterly-earnings-of-any-company-ever/>.

"Apple Earnings: 74.5M iPhones Sold, Record-Breaking Profit". GigaOm. 27 Jan 2015. Web. 28 Jan 2015. <https://gigaom.com/2015/01/27/apple-earnings-jan-2015/>

"Global Apple iPad Sales from 3rd Quarter 2010 to 1st Quarter 2015 (in million units)." Statista. 27 Jan 2015. Web. 28 Jan 2015. <http://www.statista.com/statistics/269915/global-apple-ipad-sales-since-q3-2010/>

"Notes of Interest from Apple's Q1 2015 Conference Call." AppleInsider. 27 Jan 2015. Web. 28 Jan 2015. <http://appleinsider.com/articles/15/01/27/notes-of-interest-from-apples-q1-2015-conference-call>

The Great Bandwidth Cap Boondoggle

Figure 1: Current Internet Usage Report

As of Monday, the 9th of January, our family was about to blow through Mediacom's 350GB data cap with 6 days left on the billing cycle. Over the past three months, our family's data usage has averaged 249.13GB per month. Our browsing habits have not changed dramatically this month; however, I did pick up the 50th Anniversary collection of all 23 James Bond films from iTunes and have been working to download those to my local media library.

Figure 2: Past Three Months Internet Usage

What makes me chuckle a bit is that according to Mediacom's Internet Usage page, the data I have consumed to date is equivalent to watching ninety-two (92) HD movies.

How Much Bandwidth Do I Need?
According to Giga OM, Netflix indicates that 1080p HD streaming video will use 4.7GB of data per hour.  Using that figure, 341.4 GB translates to 72.64 hours of 1080p video (which is the quality I choose). Using Mediacom's 2 hour run-time example coupled with Netflix's bandwidth recommendations, then my current usage 72.64 hours translates to 36 movies - NOT 92.

As Gizmodo reports, the picture gets even worse when you consider UltraHD (i.e., 4K video) which will use 4 times the amount of bandwidth as 1080p content - 18.8GB/hour. The new H.265 codec (which is not yet found in many consumer electronics devices) will drop streaming requirements for UltraHD video to 7GB/hour. Using the more efficient codec, the 350GB bandwidth cap will afford our family the opportunity to watch 25 UltraHD movies.

Take a look at this 2012 paper entitled "Capping the Nation’s Broadband Future?" which talks about how data caps have very little to do with network congestion. While traffic on US broadband network has increased, the cost to deliver that service is declining. Consequently, "...broadband is an incredibly profitable business, particularly for cable ISPs." In addition, "[t]iered pricing and data caps have also become a cash cow for the two largest mobile providers...".

The following are Internet Connection Speed Recommendations from Netflix:

• 0.5 Mbps - Required to be broadband (under previous rules - now its 25mbps)
• 1.5 Mbps (0.675 GB/h) - Recommended broadband connection speed
• 3.0 Mbps (1.35 GB/h) - Recommended for SD quality
• 5.0 Mbps (2.25 GB/h) - Recommended for HD quality
• 25 Mbps (11.25 GB/h) - Recommended for Ultra HD quality

The Sound and Vision had a nice little infographic illustrating bandwidth requirements.  For 720p, they recommend 3.375 Mbps (1.52 GB/hour). For 1080p, they recommend 6.75 Mbps (3.04 GB/hour).
Mediacom's estimate on the number of movies that one can watch with 354.2 GB of bandwidth is 95. This equates to 1.86 GB/hour or 4.133 Mbps. Per the Netflix web site:

"HD availability is subject to your Internet service and device capabilities. Not all content is available in HD. A download speed of at least 5.0 Mbps per stream is recommended to receive HD content (defined as 720p or better), and you'll want to make sure the Netflix plan you're currently on allows streaming in HD. To see which plan you're on, visit the Change Streaming Plan page."

So Mediacom's estimate of 4.133 Mbps for HD quality is below what the leading purveyor of streaming media recommends. Even at those minimum rates, we are talking about 720p as a MINIMUM - not 1080p. Netflix stating that streaming 1080p video consumes 4.7 GB/hour equates to a 10.44 Mbps connection speed.

Mediacom's estimates for the number of HD movies one can watch with the bandwidth cap allocation is, in my opinion, overstated.

Exceeding the Additional Limit
By today  four days later, I've gone through another 50GB. In that time, I've downloaded the following movies from my iTunes account to my local computer. The only new movie that I purchase was John Wick - that came out digitally this morning at 12:00 AM.

• The Amazing Spider-Man 2 ($17.99)
• Jack Ryan: Shadow Recruit ($9.99)
• John Wick ($14.99)
• The Lego Movie ($17.99)
• The Lord of the Rings: The Fellowship of the Ring (Special Extended Edition) ($17.99) 
• Twilight Saga: Eclipse ($14.99)

This collection of movies did consume 45.15 GB of disk space. In addition, I did stream John Wick at 1080p (another 3.28GB). So there is no question in my mind that I used the bandwidth. But according to Mediacom's Internet Usage reports, 341.4GB is the equivalent of 92 HD movies. 399.1 GB of bandwidth is the equivalent of 107 HD movies - a difference of 15. Yet, 6 1080p HD movies consumed 45.15 GB of my bandwidth allocation.

At today's prices, buying those movies would cost me $93.94. Mediacom's policy is that once you hit your quota, every block of bandwidth over 50GB is an additional $10. So effectively, there is a 10.6% tax levied on the purchase of each movie.

FULL DISCLOSURE: I did get an automated e-mail message from Mediacom indicating that since this is the first time I've exceeded my bandwidth cap, I won't be charged for the initial overage.

Figure 3: Important Usage Notice

It is just infuriating as a consumer especially when you consider, according to analysis performed by Nate Anderson from ARS Technica, that ISP costs and current state revenues do not justify the need for bandwidth caps.

Let's Examine the ISP's Costs
Up until I made a change to my account back in August, they used to itemize the costs of the different services. Now it's buried into one package price. I'm still using the EXACT same services as I was - they're just being billed differently. When they itemized the bill, the package price for the 25Mbps download rate that I have is $49.95. This works out to be $0.14/GB. Bandwidth becomes more expensive when you go over your quota - then it's $0.20/GB. You have to push the electrons EXTRA hard when you're over your quota.

Since I don't really know what fees ISPs incur for Internet transit, I'm going to use some analysis of what it costs Netflix to stream movies across the Internet. I'll reference the "Stream This!: Netflix's Streaming Costs" article from July of 2009 published on the Streaming Media! At that time, it was estimated that Netflix paid their transit providers $0.03 per GB to relay a movie across the Internet. So the 1080p version of John Wick that I watched this afternoon (which rings in at a whopping 3.28GB) would cost 9.84 cents to convey across the Internet - in 2009.

According to the "Internet Transit Prices - Historical and Projected" article featured on the Dr. Peering web site, average transit costs in 2009 started at $9/Mbps (discounted from there based upon volume).  For 2015, Dr. Peering reports that transit pricing runs at $0.63/Mbps - a 93% reduction from what costs were in 2009. Assuming all other things remain equal, the 3.28GB HD movie that would cost 9.84 cents to convey across the Internet in 2009 would cost 0.69 cents today or .21 cents/GB.

Using this as an example, let's consider my monthly allocation of bandwidth of 350GB. We'll assume that ALL that data comes from outside my ISP's network and is NOT available from any peering partners so Internet transit, the most expensive form of bandwidth, must be used to get it on my behalf. 350 GB multiplied by $0.0021 per gigabyte would cost the ISP $0.735 in transit costs.

Assuming I use my full allocation of bandwidth each and every month, the $49.95/month I pay for my high speed Internet service would leave $49.22 to cover ALL other overhead associated with providing my service and profit which the ISP does deserve to earn.

Mediacom charges $10/50GB in excess of your data cap. Assuming the cost basis is accurate, that 50GB of excess bandwidth costs Mediacom $0.105 in Transit fees leaving - $9.895 to cover profit and overhead.

Seems like it's good to be in the last-mile ISP business!