Recently in bandwidth Category

I'm in Atlanta speaking at the Microcorp One-on-One event about Trends in 2010. The three trends that I see for agents are the following: Applications, Quality of Service (QOS), and Mobile Broadband (MBB). But they are kind of inter-dependent. Ubiquious broadband leads to innovative uses and applications. Applications like on smartphones lead to a greater need for mobile broadband networks.

Mobile Broadband is growing. Smartphones are replacing cellular handsets. Social networks are moving to mobile devices so people can Facebook and Tweet. RIM's Blackberry brought us mobile email, but it is a standard on many phones now. Netbooks and data cards are presenting the US cellular companies with some fits. They like the additional revenue, but have to keep dropping billions on the network backhaul and capacity upgrades. (And another $45B+ on the upgrade to LTE/4G).

All this means that there are new uses for the mobile broadband, like the Kindle. Sprint's Wispernet allows Amazon to instantly download books, magazines, newspapers and blogs to Kindle devices. Machine-to-machine devices can utilize the cellular data network to provide connectivity for ATM machines, security cameras, and a host of other devices that need to communicate with a NOC or remote server.

All of this is a cycle of applications driving network usage. Ubiquious broadband driving more apps. It's one reason that the FCC needs to maintain open network and Net Neutrality guidelines in place.

Applications - like email, databases, office suites, CRM - are creating a demand for managed services, such as an outsourced IT department. In addition, businesses are looking at the Cloud - moving applications to a data center for redundancy, security, and availability - as a way to save money and stop worrying about the IT department. With applications being delivered in the Cloud or by way of SAAS or even Virtualization, Agents have a chance to offer more than just Internet Access or WAN circuits, like private line. Agents can sell Layer 2 to Layer 7 - pipe to apps. It's a way to get deeper into accounts. It's a way to offer a complete solution. It's a way to deliver on the label of Trusted Advisor.

Applications are driving sales. Voice and email are just the primary apps. Business critical data is also driving mobile broadband. Ubiquious broadband is allowing for innovative ways of accessing data. The problem becomes reliable access to the data. That's where Quality of Service comes in. QOS on the WAN is what is needed to access data reliably and quickly. The MPLS trigger is the Class of Service reliability and prioritization of data over the network. This is paramount for businesses running a truly converged network with video, database, VoIP, email and Internet riding the same pipes. WAN Optimization is selling due to the cost containment and the performance enhancement. Big bang for the buck.

So the agents can sell mobile broadband, applications via Virtualization or SAAS, and add QOS to the WAN to provide reliable access to these business critical data.

My buddy, Derek Thompson, just started at Fishnet Security. Besides peddling Bluecoat and SonicWall Managed Services, he is also selling WAN Optimization. While I have heard of this, I had no idea what it was or how it worked. So Derek invited me to sit down with Doug Kruger of Riverbed. Doug explained it simply enough.

One reason companies consider WAN optimization is if the company needs more bandwidth. It may not need more bandwidth, it may just need to better utilize the Internet Access it already has.

Multi-location companies sometimes have servers at each branch instead of consolidating the servers in one data center. This saves money on maintenance and storage, but in some cases may mean a bigger pipe to access the servers in real-time from the data center.

By consolidating servers into one location can save on manpower, maintenance, and data storage/back-up costs. This is also the sales trigger for Virtualization and Cloud Computing. Save on hardware, labor and storage.

TCP and latency are other reasons that companies buy bigger pipe, when perhaps they could just optimize the current WAN connections (or change to MPLS pipes instead of DIA or dedicated Internet access circuits at each location). TCP is not the most efficient protocol, but it works, just sometimes creating excess packet traffic across the WAN.  Latency for real-time applications is also a real problem across the Internet. Lastly, many applications, including Microsoft Office, create excess traffic on the WAN when any document is being opened. 

One of the ways that companies like Riverbed optimize the WAN is to eliminate duplicate data traffic. It's called De-Dupe and the effect is to eliminate up to 80% of traffic across the WAN. WANO technology usually will work on eliminating all the excess packet traffic on the WAN, which will usually result in a bandwidth savings. 

Riverbed tweeks TCP headers to modify the TCP window sizing issue. The tech here gets beyond me and my discussion, but by modifying the TCP packet headers much of the TCP "noise" on the WAN is cut down. By diminishing the excess packet traffic from apps and TCP, WAN optimization technology is able to save a lot of bandwidth. 

Riverbed also works on application latency in a similar vein, by the mitigation of excess packet traffic from apps like Microcosft Office. The technology guesses what the application will ask for and deliver it all at once instead of in 20 different packet streams. Many Fortune 5000 companies use WAN Optimization due mainly to cost cutting force on them by the economy. Data is business critical, so any way that you can save money on the WAN is a good thing.

Because broadband is so slow in the states, Amazon has resorted to allowing companies to mail in their portable devices for data storage and upload. According to Cloudave, it could take "82 days (with 1.54 Mbps T1 connection) to migrate one terabyte of data with 80% network utilization."  So "Amazon has resorted to old fashioned approach from the previous eras. Users can mail Amazon their data in one of the supported portable storage devices and Amazon will transfer it to users' AWS account from within Amazon's high speed internal network."  Sad really.

I'm seeing a lot of news in our space but not enough time to cover it all or analyze it, so here's just the headlines:

DPI (deep packet inspection) by cable being investigated by Congress. It scares the crap out of Boucher (ARS). Cox, Comcast, NebuAd  = new privacy law being debated (NYTimes).

Broadband download caps: in the news all week because apparently TWC said that without caps, they won't upgrade any more. Well, I have news for them: if they don't upgrade they will lose customers. Can you say FiOS, WiMAX, U-Verse, and now Wildblue is testing 18MB service.  ARS notes there are caps even when not explicit like TWC.  VZW and others have usage limits built into the acceptable usage policy.

Clearwire is being sued - class action status - for ETF (early termination fees) and network quality issues (can you say: false advertising on network performance?). (see here and my twitter pal @morisy).

And speaking of Caps (no, not hockey, how about Comcast battling it out with the former FCC chief's ruling that cable companies can only have a maximum of 30% of the entire market? If we applied that to telecom - and why shouldn't we? - we would have to break up Ma and Pa Bell (Verizon and AT&T). Please note: I am all for that.  Meanwhile Comcast's defense is Freedom of Speech.

Lastly, Facebook exec becomes new CEO at MySpace. Too little, too late? And Yahoo! is closing down GeoCities free hosting services, which it bought in 1999 for $3.5B. The analysis of the deal is on Fred Wilson's blog. Worthwhile read for start-ups about what VC deals look like.

So a client calls me. An agent sent him an unsolicited quote for 10MB of bandwidth over Ethernet for a ridiculously low quote. Since I have a long history with the client, he offered me the chance to match it and take the order.

So I go back to the carrier and ask for a match. I'm told no. "A competitive quote on the competitor's letterhead is required, which must be 10% to 15% less than the discounted price available with this promotion."

There are a couple of things wrong with this:

1. Why can't I offer the same quote as the other agent -- it's the same carrier?
2. How did tthe other agent get this price without a competitive quote?
3. Why do you want me shopping this client around? If I found a better quote, I would sell him that because you made me jump through these ridiculous loops!

This is the type of policy that drives agents crazy. (Where's the common sense?)  It's one reason that resellers are taking business from the carriers.

L3 released 4Q08 numbers last week. Remember at Christmas, rumors swirled about a possible bankruptcy. Now it rings up a profit.

Level 3 Communications has recorded its first quarterly profit in six years for the fourth quarter of 2008. The operator reported a $44M profit for the three months to 31 December; although it also reported lower revenue at $1.05 billion for the quarter than it did for the same period in 2007. Level 3 had a $290M net loss for the full year but this was considerably less than the $1.1B loss incurred in 2007. Full year 2008 revenue was $4.3B [about the same as 2007][.telegeographyg]

Even when you hit some numbers, The Street kicks your ass. But when you say, "For 2009, the company said it expects continued revenue weakness over the short term but is also working to cut costs" in Forbes, your stock is taking a hit.

Full disclosure: I rep for 20+ carriers and Level3 happens to be my biggest carrier. Pricing pressures are driving rates down on everything from Transit to Transport. Not everywhere mind you, but in the top 10 Metro areas, where L3 competes with Cogent, HE, and NTT/Verio, it is pressured to lower rates.

XO competes with it in many markets because XO has IRU's on L3's cable, so about 60% of where L3 is, there's XO. Surprisedly, pricing pressure is also coming from Qwest on longhaul routes, especially on Waves. (I rep for XO and Qwest as well, but not Cogent, HE or Verio). But the real competition comes from its own Resellers, like WBS Connect. Lately, Scott's company has gone underwater to take a deal from me, I mean, from L3 directly. I don't know where all this ends but I do expect HE, Cogent, WBS and others to start seeing more bad debt and late payments. The people who fight for price the most are also usually (not always) the ones who pay late. We'll keep watching.

Rocky Mountain News reports that Level3 "remains unprofitable, mired in more than $6 billion in debt, and shares have fallen to less than 70 cents. Ordinarily, Level 3 stock would be delisted, but Nasdaq has suspended the $1 minimum requirement until April 9 because of the country's economic turmoil."  With revenues flat, L3 is looking to give itself time to repay all of its debt to forego bankruptcy.

Fox and MarketWatch reported that "Level 3 Communications Inc. had its corporate credit rating cut to SD, or selective default, from CC by Standard & Poor's. S&P also lowered ratings on Level 3's convertible subordinated notes due 2010 to D from C."  In addition, Level3 changed its stock buy back plan -- and is having trouble getting takers. According to Denver Business Journal, "Level 3 is trying to push out repayment of about $1.14 billion of investor-held debt coming due in the next two years -- $305 million in 2009, $836 million in 2010 ­-- to preserve its liquidity in the face of flattening revenue. .... The company had committed to spend as much as $438 million of its cash to buy back some of the 2009 and 2010 notes. It planned to fund the rest of the debt buyback by selling debt notes that mature in 2013 and yield 15 percent interest annually."

 

According to the WSJ, " The federal government's economic stimulus package will include investment in broadband Internet infrastructure and funds to upgrade and repair the national power grid alongside more traditional funding for road and bridge repair, a senior aide to House Speaker Nancy Pelosi said Tuesday."

It's much needed, since Clearwire/XOHM has decided to slow down (cough!*) its deployment schedule because the $3.2B in cash from Google, Intel, and MSO's, just is not enough to roll out their whole network. Clearwire needs $2B more.

And apparently, VZ only wants to over-build its affluent DSL areas with FiOS. It ignores or sells off rural routes.

TWC and AT&T are moving to tiered broadband pricing plans. Heck, TWC is even raising its rates on stand-alone Internet. TWC is also vulnerable to customer switching. In recent bundles from former BellSouth, I did not see a DSL price under $30. What happened to Naked DSL at $20 and DSL Lite at $10 (merger conditions)?

The next ones to ask for a bail-out will be the wholesale Internet backbone companies like Level3, Cogent, Global Crossing, Qwest and Savvis, according to this report on Circle ID.

According to the CWA (telecom union), it's all about new jobs.

Mr. Cohen said 100,000 jobs could be created by immediately investing in more high-speed Internet networks across the country. "Jobs is the best single stimulus," he said.  The demand for services created by broadband Internet access could create another two million jobs, Mr. Cohen argued.

[Editorial note: * As if Clearwire had an aggressive roll out strategy. $3.2B should be enough IF you actual bring in profitable revenue per tower. I know, not a usual business plan for public internet companies.]

"Nemertes Research continued to throw cold water on the future of the Internet last week, releasing a study projecting that demand for bandwidth on the Web would exceed its capacity by 2012." [PCWorld.com]

I just don't see how that can be. With 40G pipes starting to emerge. With Content Delivery Networks (CDN's) popping up every where.  The CDN's make most traffic local.  The CDN's are getting closer to the edge. Does that report mean that the backhaul traffic will overflow? Or the CDN network will exceed capacity?

Nemertes analyst Mike Jude says, "More and more applications are coming online that will drive expectations for service quality even higher," he said. "I'm not saying that the Internet is going to crash in 2011, but that people's expectations are going to be throttled. People will stop going to the Internet for those services." [PCWorld.com]

Jude goes on to say that people expect more reliability and real-time traffic from the Internet (which it was not designed for). And to get thta ISP's will have tiers of packages to deliver it. That's where the Net neutrality debate unhinges.

My skepticism tells me that it is just the Duopoly wanting to make as large a buck as possible from consumers to continue to get their 40% margins, despite the fact that their actions stall innovation and the economy. And the capacity can be made available, it is just more expedient to create a supply issue.

So on Linkedin, Neal Lachman, asked if the $100 Triple Play was Viable in today's economic molasses. Neal writes:

Bundling voice, video, data services for a higher ARPU was an obvious, great move when broadband services and advanced digital services were first introducded......  However, the market is moving more towards a lower ARPU for the triple play services. This is especially going to play a big role in future operations. The time of high ARPUs is going, and soon it will be history.

I believe operators have to lower their ARPU estimates from 2010 onward, simply because the customer won't be willing to pay as much. Today operators generate $100+ revenue per month on their triple play services. In 2010 and later, they should be happy if they can reach ARPU of $50. One example is the FTTH service in Holland, where people do not even want to pay more than €50 for their triple play bundle.

My thoughts on it are here:

Telcos like AT&T and Verizon are actually losing money on triple-play. Think about the fact that they were getting $35 for a phone line and $35 for DSL (averages for consumers 2 years ago). Now they have to upgrade the network to offer TV, which is the least profitable service. And do that for $30.

Install and maintain the network that they will be capping. Install home equipment like ONT and STB. To give it away for $100. Now usually the telcos will add taxes and fees on that to increase their profit.  But its the MSO's who are making out. They went from the least profitable service (TV) to the more profitable services of phone and Internet.

With all of the CAPEX for DOCSIS upgrades as well as FTTx and WiMax build-outs, these companies won't be able to lower ARPU for triple play.

The cost of TV content is increasing. Must carry TV channels are now asking for a bite of the pie. You have seen the battle that NFL Network and the other sports networks are having to get carried by the systems -- and to be carried in the most popular packages.

I can see how the MSO's and telcos would have to lower ARPU averages in the face of the economic tsunami we are experiencing, but they won't be offering triple play for $50.

Remember that for the Bells, RGU's include security, cellular, and now tech support. Cablevision rolled out a $350M wi-fi network in NY. The duopoly knows that to keep churn down, they have to get sticky with ubiquitous Internet Access and to get close to a quad-play. Surprisingly, while Verizon has the quad play in my town (Tampa) - FiOS TV, Internet, phone and Cell - that is not the package that they advertise to my house Every Single Day.

 The cost of customer acquisition, retention, advertising, tech support, customer care, bad debt, security, upgrades, and maintenance are too high for the triple play ARPU to drop below $99.