April 21, 2009

The Sun/Oracle deal is all about databases, not hardware

When IBM first put forth its bid for Sun, it didn’t make a lot of sense to me – too much redundancy and a case of the whole not really being greater than the sum of the parts. Not so for a combination with Oracle, which made its $7.4B offer for Sun Micro on Tuesday. (Interestingly, in reviewing the blogosphere, kudos go out to an article from Neil McAllister which accurately predicted Oracle as the most likely buyer for Sun).

But the reporting to date on the deal strikes me as missing the strategic value of Sun to Oracle.

There’s two important points to make here. One that is being talked about too much and one that needs a lot more focus.

The first, regarding the  hardware/software integration angle is valuable for Oracle from an implementation perspective but not especially strategic. For Oracle, and Sun’s ability to build servers mixed with a rock solid operating system and virtualization is reflective of the growing convergence of the operating system, applications, hardware, and networking. Cisco’s increasing concentration on software and building blade servers is a perfect example of this land grab within the enterprise and the data center. Over the coming five years, we’ll see companies like IBM, Microsoft, Cisco, SAP, HP, Dell, and Oracle all looking to be able broaden out their offerings. Why?

What we’re seeing is a move to turnkey systems, especially now that virtualization has moved into the mainstream. With applications like Vmotion from VMWare, IT people can move running applications between server machines, providing an ability to consolidate multiple applications onto a single hardware platform. In this environment, does it make sense to just sell the server software (Microsoft) or the database software (Oracle)? If HP is going to build Ethernet switches to Cisco’s chagrin, then Cisco is going to push back by selling blade servers. For the end customer, this development is great – and should result in easier to install and maintain IT infrastructure, at a lower total cost of ownership. For companies without a full spectrum of products, it gets harder to compete because of bundling – for example, IBM can take a haircut on hardware in order to get long-term service contracts. Could we see a consolidation down to just a few mega-companies over the coming years in the enterprise and data center market? This could be quite likely.

With Sun products integrated with Oracle applications, we could conceivably see an entire server ecoosystem pre-configured and ready to go out the box, and since the applications need to be locked to a particular hardware configuration, this will put Oracle support engineers in-front of customers from cradle to grave, which is important no doubt.

The more interesting strategic consideration here is on the subject of databases. One thing that struck me as gravely wrong is conjecture that Oracle will de-emphasize Sun’s open source database called MySQL, which is extremely popular in the open source arena, in order to reduce competition for its core database product. In contrast, Oracle’s bread-and-butter has been its Oracle database, which is the gold standard for high-performance databases in data centers and enterprises. Would Oracle shut down this free database in order to stop it from cannibalizing high-priced Oracle sales?

This seems highly unlikely. Due to the nature of the open source community, killing or abandoning MySQL isn’t really possible, as it really has a life of its own – introducing controversial changes to an open source project simply means that the project gets forked off, allowing people disputing the change to start their own project under a different project umbrella. There are other free database alternatives, such as PostgresSQL, which could increase in popularity if Oracle somehow manages to torpedo MySQL.

On the contrary, I think that MySQL is really the crown jewel here for Oracle and a product that could fit well within Oracle’s product line-up.

MySQL = Oracle Light. Currently, developers need to create applications around either MySQL or the more powerful Oracle as there are differences in the way that developers can interact with the database language-wise. One thing that Oracle should do with MySQL is relly make it Oracle-lite. One way of doing this would be allow applications currently using MySQL to migrate up to the full Oracle package when they outgrow a free database application.

This compatibility becomes critical as applications increase in size and more importantly as server applications move into the data center, aided by virtualization and software-as-a-service. As more applications get run on a single server, I could see the data center operator starting to choke a bit using MySQL. However, this provides a perfect opportunity for Oracle to upsell data center operators – if Oracle can make its Oracle database fairly compatbile with MySQL, then a data center could consolidate down to a single Oracle high-performance database instead of relying on the free alternative.

At the end of the day, Sun’s MySQL database (acquired for around $1B a few years back) is very interesting strategically for Oracle because it helps Oracle build a presence in the low-end of the database market, and puts Oracle directly in-front of customers that are  looking to get more performance or reliability from their existing database solution. For IBM, the value of buying Sun increases – buying JAVA simply to keep it out of Oracle’s hands might be…  priceless.

The bottom line here is that this deal appears  to be very good for Oracle in the longer-term, a might be bad for companies like HP/Dell/IBM because Oracle building stronger relationships directly with the end customer. And for companies without the full gamut of products, things will be getting tougher as sales forces from these large integrated companies will not only push their in-house solutions top to bottom but will also offer bundle pricing galore.

For Microsoft, this deal is bad as well, as the company has seen strong growth of its Windows-server based database product, which is lower priced on a unit basis compared with IBM and Oracle database offerings (though Microsoft does this because it’s margin on the operating system itself is juicy). With Oracle being able to push a free database model (combined with professional services), this could help take share from the Redmond giant at the low end over the coming years. In summary, as with the case of Cisco, which builds tangential businesses which ultimately help to sell more routers,  this deal for Oracle could very well help to sell more Oracle databases down the line, creating a “minor league training facility” so-to-speak.

Leave a Comment

Filed under News

Tags: , , , ,

April 2, 2009

Does RIM’s App World measure up to Apple’s App Store?

RIM’s App World officially opened up for business yesterday, and the big question on everyone’s mind is “How will this measure up versus Apple’s iPhone App Store?” Ultimately, this will be determined not only by end user adoption of applications served up by the store but also by the number of developers providing applications as well. For RIM, this is a much needed capability, and up until now, the market for add-on applications for BlackBerries has been a patchwork composed of carrier mobile stores, RIM’s own site, software company sites, and third-party web sites like Handango. Getting customers to buy these apps has understandably been somewhat difficult, as each store had completely different ways of paying and of getting the apps licensed. In contrast, Apple has gone for simplicity, handling the licensing and payments (tied to a user’s Apple ID which is connected to a credit card) and sold around 300 million apps through the end of 2008.

The trojan horse for Apple was also application support for the iPod Touch, and Apple has said that it’s sold more than 30 million iPhones + iPod Touch devices in total. How does this compare with RIM’s opportunity here? While RIM has shipped more than 25 million devices through 2008, applications on App World can only be for relatively new phones (ones with the scroll ball or the Storm screen click). With Blackberry subscribers around 14 million, most of these users should be on a Pearl or better — a near-term opportunity of 15 million would be a reasonable guess here. An audience of 15 million users is nothing to sneeze at in other words.

How will App World measure up for developers? The revenue split is better than Apple’s (80/20 vs. 70/30 for Apple), though the cost differential might be negligeable at the end of the day. This system might work better for larger software companies… here’s why.

The payment system being used is E-Bay’s Paypal, which takes a cut of 2.9% for payments out. This impacts developers of all size. The other pricing detail is that paid apps need to cost at least $2.99 (compared with $0.99 for Apple’s App Store), though this is subject to revision in the future. Submitting applications to the BB App World will cost USD$200, and this includes 10 applications submissions. For small-time and part-time developers, this might be a big deal, as registering as an iPhone developer costs $99 all-in. This will dissuade users from creating Freeware applications, because developing multiple updates to multiple applications could end up costing thousands of dollars over time. Larger software companies won’t have a problem with this, as the superior cut of profits will pay for the higher application submission costs.

This difference in pricing for developers and consumers suggests that RIM is trying to go up-market somewhat, dissuading applications that are extremely trivial in nature. Does this make BB App World less fun? Perhaps. It’s the low-end nature of the iPhone App Store that allowed an application like Koei Pond (a best-seller on the App Store) which simply has fish swimming in a pond, with the user able to interact with the fish using the touchscreen.

There are some advantages and disadvantages to RIM’s BB App Store, based on initial observations:

1) Licensing is more flexible. While the initial incarnation of BB App World is only allowing the static model (no license is required when the app is purchased), the developer interface does allow for pool licenses (multiple licensed purchased at one time) and dynamic licensing (a license key is generated at the time of purchase). The dynamic licensing could come into play with users purchasing trial version of the software – it’s possible that software companies may want to simply provide trial versions of their applications and get 100% of the license money later on by handling their own payment system. The other way of getting around this revenue split is to charge for other services – for example, an electronic fax software could be free but charge per fax delivered.

2) Managing the carriers. On one hand, the carriers have to be unhappy with this move from RIM, because it means that carriers are one step closer to becoming the “dumb pipe”. However, unlike the iPhone App Store (which can also be accessed over Wi-Fi), the BB App Store will control how apps are sold on different carrier networks. The developer kit does indicate that application developers can restrict where the app gets sold, and this also means that carriers will have the ability to restrict whether an app gets put onto their network. In the future, hypothetically speaking, Rogers Wireless could restrict a Skype voice application from being distributed on BlackBerries in Canada. Good for the carrier though not so good for the consumer.

3) Fewer conflicts… maybe. The big advantage for RIM in all of this is that RIM has fewer conflicts than Apple in approving new software applications because it really doesn’t do software very well. In other words, RIM won’t be putting out a next-generation converged communications app (for example) and developers won’t have to worry that it will have worked on something for months only to have it rejected by Apple. For example, Skype on the iPhone must have been in the works for a long time and is now only seeing the light of day. As noted above however, the big worry for developers will be on how much input carriers have on how an app gets distributed.

One important observation is that RIM appears to be targeting more professional developers with their App World based on the higher up-front charges for getting the app on the store. Where are these fees going? While some of this cost may be going to more testing expenses (there are a lot more models of BB that need to be supported… at least 7 families of devices). Is the Paypal integration a downside for the BB App World? Perhaps yes. I would have liked to have seen RIM manage the registration process themselves, getting closer to the end customer so to speak. They’re already setting up a registration process called MyWorld which will allow users to re-download purchased applications, so perhaps RIM is simply getting more payments sources lined up (Google Checkout perhaps?) Even better, this was the perfect time for RIM to set itself up as an intermediary between the carrier and the user – allowing users to pay for apps using the wireless account or prepaid accounts. In my view, RIM needs to add more payment options at this point, going for direct credit card payments, debit cards, etc in order to make the process as easy as possible for users.

Who benefits from this payment system in the BB App World? Aside from eBay, another undisclosed beneficiary of RIM’s payment system is e-commerce company Digital River (NASDAQ:DRIV). According to the App World vendor guide, paid applications going onto the BlackBerry app site, developers will need to employ technology from Digital River to handle payments through PayPay, and a percentage of that 20% cut for RIM goes to DRIV.

In summary, the BlackBerry App World is an important for RIM as it signals that it recognizes that applications will be an important competitive point differenting smart phones goings forward. Thus far, my opinion is that the App World will be different from Apple’s App Store, appealing to more upscale developers based on the difference in pricing schemes. This could by extension result in a small developer community as well. While RIM may ultimately miss out on some innovative applications (something which has been going on from the beginning), the upside is that we should expect hundreds of valuable new applications to show up over the coming year to RIM’s site, with a SIMPLER payment and licensing system than before. The bottom line for RIM is that as a developer, I think the new pricing and licensing scheme is workable. An improved system should result in better traction

The only downsides in terms of supporting both the iPhone and BB ecosystems remains the difference in programming languages, and if I’m writing an app in C++ for the iPhone and Java for the BlackBerry, the actual development effort is nearly doubled. On top of that, while the iPhone has a standard hardware set to deal with, the BlackBerry ecosystem still has important difference and capabilities to deal with. For example, the Storm has a touchscreen and has a higher resolution (480×360) than my current BlackBerry 8800 (320×240), meaning that BB app developers will need to tailor their software for working on different devices. Ultimately, the big question is whether developers will flock to the BB App World and I think the answer is “In time”. Especially for low-end developers, the resources in terms of money and time and addressable market just don’t add up to release an app on BlackBerry first before the iPhone.

The key takeaway for me is that I think RIM took the safe approach in designing its pricing scheme, ensuring that costs incurred in evaluating new applications for its app store were paid for up-front by developers instead of getting its money back through a cut of revenues from the apps. It’s a big improvement no doubt, but will it be enough to shake up the developer landscape at a time when Nokia, Google, Microsoft, Palm, and Apple are all vying for developers? I’m not so sure.

Leave a Comment

Filed under News

Tags: , , , ,

March 24, 2009

Microsoft finally embraces open source apps in the server realm to help fend off VMWare

At last week’s MIX09 web conference, Microsoft made a couple of important announcements.

The first is regarding Silverlight 3, the latest version of Microsoft’s web application environment (competing with Adobe’s Flash and Air products). The key new feature for SL3 is an “out of browser mode”, which enables applications written using the system to not only be run from within a regular web browser – they can be downloaded to a user’s computer and then executed independently. Why is this an important new capability? With the mobile world seemingly gaining a lot of new focus from app developers (iPhone, Android, Windows Mobile, BlackBerry App Store, Palm WebOS, Nokia Ovi) and some netbook makers starting to look to bring these environments to larger form factors, the battle for app developers is really on now. If SL3 is brought to the different incarnations of Windows 7 for desktops, servers, and mobile devices, it would provide a way for developers to write applications once and have them usable on all of these platforms. The disadvantage for regular client applications (like Microsoft Office) is that these programs aren’t easily transferable between systems (they would need to be adapted and re-compiled) but SL3 apps are runnable on any platform, and now using this out of browser mode, can even escape the confines of the web browser and look like a regular application.

Google Apps meanwhile has been trying to escape the confines of regular browsers as well, and through Google Gears can now save data to the local computer (for running in an off-line mode). Moreover, with Google’s Chrome web browser, Google applications can be launched from a computer as if it were a standalone application. Offering the advantages of seamless document sharing, plenty of on-line storage, being free for regular consumers, and integrated with a web site domain for enterprise users, Google Apps using the Chrome web browser becomes a formidable threat to Microsoft’s productivity money-making machine.

google-apps

I also took Adobe’s AIR environment, and an application called TextFlow in particular. TextFlow is a program that allows users to share documents with others, and then allows an editor to review and merge changes to produce a final version. Downloading the application was a seamless process, and the Installation took place all within a matter of minutes directly from within a browser (The Adobe Air environment was installed first). Registering TextFlow only took a few moments through an e-mail check… the end result? Absolutely terrible. I had a chance to look at Adobe Air applications last year and thought that while the user interface was stunning, I wondered who would want to use this? The view is pretty much unchanged, and using the TextFlow non-commercial version, a document I uploaded not only wasn’t really editable at all using a graphical interface to control formatting (unlike what I saw in the demo video), but worse, the document I uploaded disappeared completely and trying to re-upload a simple Word file caused the program to hang up. The bigger problem for Adobe Air is that while Adobe Flash is installed in almost every web browser in the world these days, Adobe doesn’t really have access to users – it’s missing a platform…. unlike Facebook, Google, Microsoft, and Yahoo. Through Google Apps, the most important thing that Google gains is access to users that are interested in enterprise apps, and this is something that Adobe simply doesn’t have at the moment. Should Adobe really be buying Yahoo in this context? Hmm…

The second important announcement for Microsoft was the beta launch of the Microsoft Web Platform. Why might this be an unprecedented time for Microsoft? For the first time, we’re seeing Microsoft really embrace open source applications, which by their nature are and will remain free. Microsoft has been milking enterprises for a long time now through its dominance in the space, and charges a hefty price for even its most basic server offering (around $1k for 5 users, $2k for 15 users) When was the last time that Microsoft really gave something away for free? For companies like Real and Netscape, the answer might be “only when they want to torpedo another company”. With Microsoft’s Web Platform, we see a fundamental change in the approach Microsoft is using, adopting more of a Google model where they’re trying to add value for customers using their ecosystem. According to an eWeek article, the product manager for this product, Lauren Cooney, comes from IBM and BEA and brings a different attitude towards open source applications and emerging technologies not built from within (“Make Web not war”).

This attitude shift is something that I’ve been a proponent of for a long time. After all, companies these days may be running a mix of different server technologies and applications, and want to be able support all of them through a minimum of re-engineering work. Through the new Web platform, we see Microsoft supporting not only its own ASP.Net technology, but adding support for a free programming environment called PHP out of the box. Even better, Microsoft will now offer important new open source packages through their enterprise app store for free, including popular packages like WordPress and Gallery, enabling hosting companies and enterprises to add these capabilities on Microsoft architectures, and I’d expect other major open source tools like Joomla to be on this platform before long. It’s not perfect yet though, and where the buck stops might be in offering Sun’s MySQL database (perhaps soon owned by IBM). After all, MySQL competes directly with Microsoft SQL Server… the problem lies in the fact that a lot of open source packages use the open source MySQL as a fundamental database – would Microsoft force new open source packages to allow them to interface with Microsoft’s database? No, but I’d expect that Microsoft would want to plug this hole by at least submitting new open source changes to these packages to at least allow Microsoft to not have to also install MySQL at the same time as they install these new packages.

What does this have to do with virtualization? Perhaps a lot. As a consulting client ponders a new server architecture, the fundamental question arises. Go with a costly Microsoft server architecture or a free Linux server? One reason to go the Microsoft route would be to enable legacy applications like payroll and accounting software to be seamlessly transitioned over to the new hardware. Finding alternative software packages for these applications would be time-consuming and risky. Up until a few years ago, companies had no choice but to buckle to Microsoft’s demands and buy new licenses. However, with virtualization, this suddenly opens up the possibilities. For these old applications running on Windows Server 2003, users can now use tools from companies like Microsoft and Citrix to virtualize these old environments, essentially simulating them either on bare metal (no operating system) or within Linux. Ironically, Microsoft moving into the virtualization market with its low-cost Hyper-V offering forced VMWare to go with a free product (ESXi), and VMWare hopes to make money on it through support and training. For a lot of small businesses, this could be a great and free way of supporting old Microsoft server installations (which are already paid for) in a Linux environment. Better yet, it turns out that the risk might even be lower for these business because they don’t need to upgrade to a new operating system in Windows 7 server, which might have forced them to upgrade their application software at the same time – killing two birds with one stone so to speak. The other option that Microsoft was hoping for was to get customers to buy the latest Windows 7 server, and for applications that they want to consolidate onto the Windows server, they could use the Hyper-V virtualization, while also enabling free Web applications as well. Does this strategy work for Microsoft? No, but at least it narrows the price gap, making it easier for companies that really want to stay with Microsoft to justify away from Linux. For example, companies wanting to set up a corporate blog might now simply install WordPress on a Windows server instead of buying a $1000 Linux box to run the popular blogging software.

In summary, Microsoft in my view appears to be moving in the right direction in the server area, and Microsoft Exchange (supporting Outlook) is still a dominant technology in the enterprise. And while virtualization is still a big threat to consolidate Windows applications on Linux platforms, Microsoft’s apparent embracing and promotion of open source applications does signal an important change in Microsoft’s “only built from within” mentality. How important is the server area for Microsoft? With an operating margin of almost 40%, it was responsible for about 25% of the operating profit in Microsoft most recent quarter. Definitely an important piece of the puzzle. More importantly, keeping companies “Microsoft only” is definitely a starting point for Microsoft’s Office franchise, which is even more important to Microsoft. What would happen if a company doesn’t use Exchange and instead goes to IBM’s Lotus suite or Google Apps? While Word remains the dominant word processor in the corporate world, it could lead to more usage of free/low-cost tools like Google Docs. For example, in writing this blog, I use a simple text editor and then on-line WordPress.com HTML editor for formatting, which is sufficient when you want the content to be viewable on a number of screens, including within a mobile context. The important jewel for Microsoft is its business tools division, which earned 53% of the company’s operating profit last quarter, with a 64% operating margin! Clearly, lots of low-hanging fruit for the competition, and with Microsoft now trading at around $18 (up 20% from its low at the beginning of March), it still trades at under 10x P/E.

The bottom line for Microsoft is that while Microsoft should have been more proactive in the past with regards to open source software, I’m encouraged by the announcements from MIX09 in this regard, and it does signify that Microsoft is working to improve the value proposition for its server customers at a time when they have more choices than ever before. There are implementation risks with this strategy, and Microsoft faces internal struggles as it starts to offer external software packages that internal groups might have been working on (such as MySQL vs. SQL Server). At this low relative valuation for Microsoft, I do see that Microsoft will be incrementally more successful with this strategy going forward in protecting the “family jewels”.

Leave a Comment

Filed under News

Tags: , , , , , , , , ,

March 20, 2009

Cisco attempting to make Internet video work right – Putting the Pure Digital acquisition in context.

The time has come for Cisco to make a huge play in the home.” – John Chambers, Jan 2009

Users want quality and they want it now. Today, the Internet experience is clearly completely different from 10 years – accesssing the Net over dial-up access was a trying experience, waiting for even the graphics on a Web page to fully load… and downloading one file might slow down surfing to a crawl. Today, the experience is much better courtesy of video services like YouTube (which uses Adobe’s Flash technology for the streaming), but real-time video remains very much a hit or miss phenomenon, and largely low quality through services like Skype Video. Want to try a video chat session over a Net Cafe’s Wi-Fi connection? Good luck.

In the enterprise, a high-end two-video solution is possible these days, but it can be prohibitively expensive, with each video room costing well north of $300k, not to mention dedicated bandwidth from service providers. Yet, the experience of Cisco Telepresence is something to behold (I’ve tried it personally twice thus far) and the exciting value proposition of very good two-way video is that you quickly forget about the technology, and it “feels” like you’re communicating in person. A smart thing that Cisco requires is that installations of Telepresence need to have precise room specs, including the room backdrop and table heights. Why? The circular table extends into the video conference wall, simulating talking with someone across the room.

tp

What does this have to do with Pure Digital, the maker of the popular Flip series of portable video recording devices? After all, it’s a big premium on a company which may do around $200M in sales this year in a relatively low-margin consumer business. In my view, this deal is all about streaming video and more importantly two-video in the long-term, and for Cisco, it appears to be very-well positioned for this transition given that it dominates in a number of areas where intelligence is required. It all starts in the core of service provider networks, and Cisco’s share among service providers is very strong, around 50-55%, starting with its largest router, the CRS-1 core router. As IP routing is being built out throughout service provider networks, Cisco is winning more than its fair share of these equipment sales, and its acquisition of Scientific Atlanta in 2005 for $6.9B brought a huge video play for the company, and a strong presence at the cable companies. The purchase of Linksys cemented Cisco desire to go “end-to-end” with a strong profile in the consumer’s home (combined with S-A’s strong share in cable digital boxes and PVRs). The end game is moving an increasing amount of video content around the home and over the Internet, and two-way video seems like it could be the next “killer app”.

The big problem in the consumer arena? There are no standards to speak of… well, that’s not quite true. There’s just a lot of them. As we’ve seen with the Blu-ray vs. HD-DVD battle of a few years ago, major CE companies really want their protocol to be the way that machines communicate because it means significant $$$ in royalties. In the convergence of CE devices such as televisions, set-top boxes, packaged media players, and computers, figuring out a way of moving information between them has been challenging, in part due to the concerns of the content companies. Wi-Fi appears to be one of the basic building blocks of most consumer networks, providing basic connectivity between devices, but beyond that, figuring out how to share files between Sony’s PS3 and a PC for example is… rocket science. (I’ve used a tool called TVersity which can provide “translate” between the two devices). Ultimately, getting devices more intereconnected needs to be completely idiot-proof. In this light, is there a role for companies that can be glue that holds in all together? Absolutely.

Turning back to why Cisco wants to own the end devices, ensuring video quality is absolutely about having knowledge about the end-to-end connection. A good example of this is in AT&T IPTV installation, where only the information between watched is sent between the AT&T’s video servers and residential neighbourhoods, helping to reduce the load on AT&T’s network. With AT&T’s network barely able to deliver 2 HD channels to each home, there’s a complex handshaking going on between AT&T’s network and the set top boxes in the consumer’s home to get channel changes working fast. The concept of the channel change might be the closest thing we have today to requiring a very fast response time for high speed video (HD video rentals are not as demanding because if a video starts a minute later, that’s ok but clearly this isn’t good enough for a video chat). Now in bringing two-way video to the consumer arena, it will be critical that Cisco help to not only develop the home networking market, but also have end devices that effectively communicate all the way through the end-to-end path. For example, will service provider networks want to know if there’s congestion taking place within a consumer’s home network? Absolutely. At the end of the day, how fast video is streamed is limited by the slowest element in the path. Likewise, if an end-device is transferring information that isn’t time sensitive, it would be helpful to flag this information as “best effort”, helping not only the service provider but also the consumer in terms of getting a better overall quality of service by not trampling over time-sensitive services like VoIP… and two-way video calls in the future.

The next step for Cisco with Pure Digital could be to gear the video recorder towards two-way video over home networks, especially since the Flip Video camera already supports a video standard called H.264, which is currently in use in a modified form in Cisco’s high-end Telepresence solution. Longer-term, it’s possible that we could see a Cisco Pure Digital Home Telepresence solution, which would allow high-quality video calls in a handheld form factor. Where this gets really interesting is if this mini-Telepresence solution could then be placed on top of a user’s flatscreen TV, and then the video could be outputted through a cable from the Flip to the TV. While this sounds like a far-off prospects, the output specs of the existing MinoHD Flip camera are pretty impressive: 16:9 widescreen, HD 720p (1280 x 720) at 30 frames. If this device is paired up with a Scientific Atlanta set-top cable box, the outgoing video could be handled by the Flip video camera, and the incoming video could be handled by the set top box. When Cisco first started talking about the idea of Telepresence in the home, the key question was always how they were going to get the outgoing video integrated into the box (cable companies want this box to be as cheap as possible because they’re subsidizing the box to a great extent). The answer might be in developing the Flip to be a high-quality “outbound video” source, able to send its video content in real-time to real-time video conferencing destinations like Skype, and even Cisco Telepresence call centers! The iPhone will even play a part in all of this, as Cisco could require that Apple allow Cisco to sell an iPhone app which enables full integration into Cisco’s consumer device initiative, perhaps even enabling the iPhone to be used as the outbound video source for home Telepresence solutions.

Networking in the home network is going to get a lot more crowded, as I wrote about in articles on the subject of femtocells (providing cellular access within the home) – with service providers wanting to put their 3G (and 4G down the line) devices into user homes to help off-land smartphone data traffic onto fixed broadband connections, we could see much better performance from smartphones within the home. Does Cisco need to be a player within this emerging market? Cisco does have an investment in UK-based ip.access, which secured an order for 7M units from AT&T and was ranked by ABI Research as #1 in the 2G picocell market (larger-ranged femtocells for areas like businesses and subways) and #2 in the femtocell market (behind Google, T-Mobile VC, and Advent VC-funded Ubiquisys), while also owning WiMAX player Navini. Over time, a cable company could want an integrated device with DOCSIS 3.0 for high-speed Internet and a WiMAX/CDMA femtocell (Comcast is an investor in Clearwire), and Cisco will want to make sure it takes advantage of this move to DOCSIS 3.0 at the cable companies add support for easier home networking (the $120M acquisition of Pure Networking in 2008 was made with this in mind).

In summary, Cisco’s long-term home strategy is still geared towards them selling more intelligent routers to service providers, but with this acquisition, I do see the wheels starting to turn in John Chambers’ head about how Cisco is going to get two-way video actually moving in the consumer arena. Looking ahead, the pieces are starting to fit together through a combination of faster home networking (802.11n) and faster pipes into the home (DOCSIS 3.0, Fiber, VDSL2, WiMAX). The more challenging part is getting the devices to talk together, and in this respect, Cisco really has the most to gain from a LOT more video being generated by the consumer just in terms of more router sales. Cellular wireless might be the Achilles’ heel for Cisco in this entire scenario, and picking up Nortel’s CDMA and LTE assets in a firesale might be the more important next step for Cisco. Now trading at around 12.8x earnings, Cisco is down on the day but has had a good mini-run to the $16 level from under $14 earlier in the month. In my view, the positive for Cisco is that they do appear to see the bigger picture, and have the pieces to not only effect a fundamental shift towards video telephony, but also materially benefit from the paradigm shift…

Leave a Comment

Filed under News

Tags: , , , , , ,

March 17, 2009

Ericsson demos a new fixed line broadband technology called “Vectorized VDSL” – What does this have to do with 4G wireless?

Ericsson is on a bit of a roll these days and is north of $9 after dropping below $7 in January. Clearly, an important element of Ericsson’s resurgence is its solid positioning in 3G and 4G wireless -  the announcement of an important role as an equipment provider in Verizon’s 4G LTE network is definitely a sign of this. However, the news out of Ericsson this morning is that they are demonstrating a new fixed line technology called “vectorized VDSL” (henceforth referred to as VVDSL), which will provide a data transfer rate of 500Mbps over a 500m distance, or roughly 100x faster than average home Internet connections in North America. Why is this an important announcement and what market is Ericsson trying to address with this new product? Let’s dig into VVDSL…

What is VVDSL? In short, telcos are delivering broadband Internet connections through a number of avenues: Clearwire in the U.S. is delivering relatively low-speed connections over wireless using WiMAX (and pre-WiMAX standards), cable operators are delivering it through coaxial TV cables using a technology called DOCSIS, and Verizon FiOS is delivering through fiber lines using passive optical networking (PON). For legacy telcos however, those that don’t go the route of laying expensive fiber optic cabling end-to-end are using DSL technologies, which re-use the old copper cabling already going to houses, apartment complexes, and businesses, though fiber is generally brought to the neighbourhood and the signal is then split off to DSL to reach the last mile. And as companies like AT&T look to compete in a broadband war as cable, fiber, and wireless companies ramp up the speeds to offer more HD video and higher Internet speeds, next-generation DSL will be critical in keeping up with the pack.

While DSL has the advantage of leveraging a huge installed base of existing cabling, the problem is that the quality of said cabling isn’t great (much worse than cable’s infrastructure), and DSL technology needs to account for crosstalk between the lines especially as the rates go up. Ericsson’s new VVDSL technology looks to bring 5x the speed of the fastest current DSL technology called VDSL2, which was standardized in the 2006 timeframe. How will this be done? The technology behind VVDSL is channel bonding, which looks to take 6 different existing copper lines, and multiplex the data stream to get 5x the throughput. Sounds easy? Well, not really. As seen in this article from back in early 2007, bonding even two channels together may only increase speeds by 25-30% if the two copper channels are too close too each other. As noted earlier, cross-talk between the lines is the problem here, and Ericsson’s VVDSL technology adds important cross-talk cancellation capabilities to the mix.

The other big problem with DSL is the distance from the central office, and once you get more than a few kms away, the speed drops off markedly. In fact, the current top end DSL technology doesn’t perform any better than the previous gen ADSL2+ beyond about a km and a half. Channel bonding seems like a natural evolution of the technology, but so far, the upgrade to two-channel bonding has been slow for AT&T, though it’s desperately needed because for some consumers, their links are maxed out as soon as you run 2 HD channels into the home along with an Internet connection (probably looking at around 20Mbps here). However, the speed equation changes dramatically with VDSL2 if distances can be brought under 0.5km, and speeds of 50Mbps per user (or 100Mbps for pair-bonded links) should be sufficient for the next 5 years or so. At the end of the day, VVDSL isn’t all that exciting in the last mile for homes, in my view, except for AT&T being able to advertise “Up to 250Mbps to the home!” When looking at broadband to buildings and businesses, that’s a completely different story altogether.

VDSL2
The real application for VVDSL is in circumstances where installing fiber cabling would be prohibitively expensive and there’s sufficient demand. Pricing in the fixed broadband access arena might be getting more challenging going forward, especially with cablecos completing their DOCSIS 3.0 upgrades (which can offer more than 300Mbps downstream and 100Mbps upstream) and Verizon upgrading their fiber offering to GPON = 2.5Gbps. Ultimately, these companies are building out their backbones with fiber cabling, but VVDSL could help in providing to the flexibility to pick and choose when fiber goes into existing areas.

Perhaps the biggest driver of a VVDSL technology isn’t better end user access but as backhaul for wireless. With more consumers moving away from land lines to wireless plans, it may not be that long before we see the same transition take place in the business world. A visit to Ericsson in Stockholm a few years ago drove this point home, as employees there had already started to use their cell phones as a primary business extension. As the move to wireless (and to smartphones) takes place in both the consumer and business arenas, the importance of backhaul (providing the connection back to the telecommunications backbone) obviously increases. Providing cheap, high speed, and low latency connections back to the Telco provider is going to be the name of the game, and this year we’ve already seen a huge move towards femtocells – mini wireless basestations that go into consumer homes, which then route wireless traffic over a consumer’s broadband connection. For the service provider, this is economical because wireless spectrum is extremely valuable and for the consumer, they can save some money and/or get better performance over their iPhones.

While it’s a question mark how fast femtocells will propagate, major service providers are definitely getting their offerings ready (the value seems to be a bit more slanted towards the service provider because carriers want to pass off the cost of the femtocell on to the consumers). For DSL-based service providers, VVDSL may be able to provide really cheap broadband access and cheap wireless backhaul, using the existing copper infrastructure. In a sense, the volume of smartphone data going over the femtocell network may not strictly need this big a pipe, but these data connections need lots of headroom because of the real-time nature of services like VoIP – i.e. as soon as connections get congested, very bad things happen to time-sensitive connections. And with video teleconferencing likely to be an important service over the coming 5 years over video terminals and smartphones, moving this data over underutilized copper lines might be the cost effective way to go for a giant like AT&T.

The bottom line for the DSL world is that with VVDSL, 500Mbps is a decent amount of throughput for companies and apartment complexes especially at a time when service providers are looking to backhaul wireless connections over fixed lines. Longer-term, AT&T might have been better off going right to fiber when Verizon decided to take the plunge, but now firmly committed to DSL, Ma Bell has to pick and choose when and where it will deploy fiber. While VVDSL won’t be standardized until later this year, this VVDSL technology is a sign that Ericsson is looking at how it can help service providers navigate the waters as fixed and wireless telecommunications standards overlap and ultimately converge towards the day when all forms of telecommunications is delivered over one high-capacity fiber optic cable going into the home.

The end game for Ericsson is winning in 4G wireless, and I see that innovating on the wireless backhaul side of the equation will help Ericsson sell the entire solution to service providers… sooner rather than later.

Leave a Comment

Filed under News

Tags: , , , , ,

March 12, 2009

Google Voice launching today – Impact on the carriers and on the enterprise

There’s a lot of chatter today about the new Google Voice service, and while the impact on the consumer space is interesting, the bigger impact for Google’s new voice service might be felt most in the enterprise space.

To sum up, Google purchased a company called GrandCentral back in 2007, and it was thought that Google was looking to add telephony elements to its portfolio of messaging products. GrandCentral provided users with a free phone number, and calling that number would then ring you on multiple real phone lines (landline, mobile, Gizmo phone), taking a message if you were away. During the beta period, users could also initiate outgoing phone calls through a ringback service (it calls one side, waits until it answers, and then does the same thing for the called party) – and it was free within North America. The Since then, the GrandCentral service hasn’t changed… until today. The new service is being rebranded as “Google Voice”, and the biggest weakness of GrandCentral as a unified platform (or “one number to rule them all”) was the lack of SMS text messaging integration. In Google Voice, a number of new features were added, like SMS text message forwarding (incoming text messages get forwarded to multiple phones), text message archiving, and transcription of voice mails to text. I haven’t had my GS account upgraded yet, so I don’t if these transcriptions are automatically forwarded to cell phones, but that would be very slick indeed.

inbox2

The first implication of this is that this suggests that Google is finally ramping out its telephony product to general masses. previously was limited to beta users of GrandCentral and therefore users that wanted to join even a year ago might not have been able to. While the initial Google Voice conversion is just for existing GrandCentral users, it’s pretty clear that the game plan is to roll it out on a broad basis. For Google, this is a positive because inserting itself into the stream of who is connected to whom, and transcribing voice mails into text does mean that Google is getting smarter. While news sources have suggested that Google won’t be using advertising to make money in this venture (perhaps by inserting voice ads into calls being launched through the service), I think that Google will get its pound of flesh in one way or the other. If GrandCentral integrates its capabilities into Gmail, that would be a good thing for Google as it would drive a lot more eyeballs to the ads there. Advertising in text messages could be another route, and inserting a small “Sponsored by Company X@TinyUrl” next to it might be a harmless compromise. The fundamental value prop for Google in my view has been in creating ads that are not annoying to consumers and therefore it might be reasonable to expect Google to proceed with caution here.

The consumer impact of this launch could be important over time. Google’s Gmail is the best service amongst the big 3 in my opinion, but at the end of Jan 2008, a hitwise.com analysis showed that Gmail was only at around 5.5% penetration, well behind Yahoo at 54% and Microsoft’s Hotmail at just over 25%. The ability to integrate telephony capabilities might be one area where Microsoft and Yahoo might not follow any time soon. The impact for the carriers might be overstated to some degree, as Google Voice appears at first glance to be complementary to real phone lines. GrandCentral’s support of the free Gizmo phones isn’t a threat – I’ve been using it on and off over the past 6 months and the voice quality is just terrible. At the end of the day, users still want to be connected wherever they are – and increasingly that’s through a mobile phone. Does it attack juicy wireless carrier profit spots like voice mail and SMS? Not really in the near-term, as users still need to have a voice mail box on every single phone line they have… just in case. It provided a further negative for carriers exposed to landlines as it does mean that current landline owners can now move to a cell phone for their portability needs and a free Google Voice phone number to be a central call routing center. For example, if someone calls my Google Voice number, I can answer the call on my cell phone but transfer the call to a regular office line in order to save wireless minutes or to get enable meeting attendees to listen to the call over a conferencing system. In the longer-term, high-margin extra feature sales for wireless carriers could be compromised – one day, could we be using an iPhone with integrated Google messaging capabilities – hooked up to Google for voice mail and presence notification instead of to AT&T? Absolutely.

The bigger impact might be in the enterprise. Google now has the tools to roll out a unified messaging platform for enterprise users, all for $50 a year. How would this work? Existing Google Apps users already have access to chat through Google Talk, which shows presence information. From there, users can already call other contacts through Google Talk, or can send them a text message through the same protocol. The neat addition that happened over the past year was the ability to text message users that are on regular cell phones (enabling a temporary SMS phone number to represent the person using the computer interface). Calls could be initiated as well through Google Talk, but it required that the other person being called was on Google Talk as well. Given that regular phone lines in the workplace still rule, this was asking too much for it to be really useful. “Unified communications” as presented by companies like Microsoft, Nortel, Avaya, and Cisco go beyond basic messaging and also support full telephony with unified voice mail. This is where GrandCentral comes in, and if Google can give out Google Voice numbers within corporations (and automatically tagged with the company’s name), then this might be an exciting proposition especially for small business owners. In fact, given the cost of a business phone line and the systems necessary to get calls routed, I wouldn’t be surprised if Google is able to charge a signficant premium for Google Apps + Voice, perhaps on the order of $100 a year. In effect, this could turn out to be almost like a virtual PBX system for corporations, except Google just provides call routing to regular cell phones. Home business owners could have their business number routed to regular phone lines (simultanously ringing a landline and a mobile phone), along with free voice mail, voice mail transcriptions, and most importantly, SMS text messages. SMS text messaging to this date has been an underutilized asset in the corporate world at a time when SMS’s popularity worldwide remains surprisingly good (As an aside, Twitter’s popularity seems to be growing and it’s no coincidence that it’s short message “limitation” is perfectly suited for SMS). Now, corporate users can now receive text messages seamlessly integrated with voice mail, and it won’t be long before the artificial boundary between text messaging and e-mail falls. In other words, Google Apps + Voice users might soon be able to see absolutely everything in their Gmail accounts – incoming and outgoing voice call logs, incoming voice mail transcriptions, IM chats, text messages, and regular e-mail all in one. This is a powerful combination.

All in all, this announcement is a real positive for Google in its efforts to unify communication across a number of different avenues. The text messaging addition to GrandCentral is a game-changer in my view, and will force unified communications vendors to re-assess how they’re selling their goods, and in this sense, Google is really bringing something innovative to the market today, though I don’t think this will have a huge impact on the stock in the near-term. In the mid-to-long-term, it’s a positive for Google in that it should help drive more traffic to Google as a centralized messaging platform for consumers (though the costs of offering free SMS and telephony service perhaps offsetting this), while also improving helping Google Apps to increase its marketability as a true alternative to Microsoft in the enterprise. Personally, my upgrade to Google Voice should come within the few days and I can’t wait.

Disclaimer: I own shares of Google at the time of publication.

Leave a Comment

Filed under News

Tags: , , , , ,

March 11, 2009

Apple’s new Shuffle – Loving the new form factor but something’s missing…

Apple is now selling its revamped iPod Shuffle, which is a low-end music playing device but has been successful despite lacking a display of any sort. The new $79 Shuffle (officially revision #3) now features a sleek all-aluminum look (silver or black) and is smaller (Apple notes that it’s now smaller than the size of a battery), with more storage space (4GB). The key addition is voice feedback (VoiceOver), and the device controls have now been moved off of the face of the Shuffle to a 3 button interface sitting on the cord for the headphones. How do the prospects look for the Shuffle?

shuffle

Apple, to its credit, has been a master at marketing and is able to sell multiple iPods to the same consumer, segmented not strictly by price but also for different applications. For example, the iPod Nano is a midsized device which can play some video content and can be used to record jogging information through a Nike+ sensor built into a pair of running shoes; the iPod Shuffle is a great device for workouts, the iPod classic has a huge hard drive for people that need to store their entire collections, and the iPod Touch leverages the iPhone app store to expand its capabilities and its large 3.5″ screen is well-suited for web browsing and playing videos.

How many iPods are we talking about? In their last fiscal quarter (Q1F09), Apple sold… 22.7M iPods (11.1M, 11.0M, and 10.6M in the three previous quarters) which totals 55.4M iPods over the past year. Yet during this time, average revenue per iPod has been slipping, falling $148 last Q from $181 the year before. This trend is a disturbing one for Apple, and while some of the price decling is due to component pricing coming way down, it does reflect the challenge in selling devices into a mature market.

The one important detail of the Shuffle is that for the first time, an iPod device will now be sold without on-board controls – it does not have the ability to work without a special pair of headphones (with integrated remote controls). For US$29 ($79 for in-ear headphones), Apple now essentially cuts off consumers from buying generic headphones for their Shuffle, though several 3rd party headphones (from companies like Shure, Sennheiser, and Etymotic) with remotes do exist though are similarly priced. For Apple, while they made $4B on iPods last quarter, their revenue from peripherals was just $382M… if they can sell 2-4M more headphones a quarter at an average of around $50, they could boost their quarterly take by $100-200M a quarter, a significant improvement this category’s sales and likely high margin stuff as well.

My thoughts on the Shuffle… While the storage capacity and capabilities of the new Shuffle are impressive, taking away the on-device navigation is a dangerous game. For Apple, removing these controls cuts down on the parts that go into the Shuffle and simplifies the manufacturing complexity – ultimately helping with the device’s margin profile. But moving the controls off device (instead of leaving them there and ADDING the on-cord controls) will by frustrating. For example, I have a 1st generation shuffle with in-ear headphones that have been tailored to fit the size of my ear canals… if I upgrade to the new shuffle, I not only get an inferior set of headphones out of the box but I won’t be able to use my old headphones with it either. As seen by negative comments left in the Apple Store, some consumers are less than impressed with the quality of Apple’s included headphones and in the case of the Shuffle having to spend as much on a replacement set of headphones as they spent on the device itself could be a source of irration. All in all, while the friendly intro video at the Apple site suggests that the removal of these controls was a way of making the device sleeker, it’s a mis-step in my view. Would something like this happen if Steve Jobs was still looking over every detail of new product introductions? Maybe not, as stories of how the iPod first came about suggest that Steve really was obsessive about making sure the navigation elements of the iPod were flawless – even down to the number of clicks necessary to get going on startup. With the new Shuffle, I now have to triple-click to go back a song? Really?

Not to overstate the issue, but the lack of on-device controls does speak to a bigger issue with Steve Jobs’ importance for Apple. When Steve was in charge, he not doubt had the final say on product decisions. With COO Tim Cook in charge, product managers will likely be left with more responsibility, especially because he’s simply the interim CEO (sort of like Jay Triano’s role with the Toronto Raptors). Does Apple have someone who can oversee product decisions, take a step back, and say “this product needs to be changed because it doesn’t work right” even though it might cost significant $$$ to change it? I don’t think so. Ultimately, this issue remains an overhang to the stock and Apple might be advised to address this issue sooner rather than later.

On the positive side, one of the important new innovations with the new Shuffle is VoiceOver, which can announce track names and playlist names to allow users to navigate the Shuffle, which doesn’t have a display of any sort. How does VoiceOver work? When the device is synced to a computer, iTunes records an audio file for each song that is copied to the device (spoken in the language of the user’s choice), and then when the iPod is asked “What song is playing?”, it simply needs to lower the currently playing track and play the associated file which announces the song’s name. Simple, low-tech, and probably patented.

The importance of VoiceOver can’t be overstated. One of the weakness with the Shuffle has been the inability to use playlists, which meant that a user had to just put all of their music on the Shuffle and then navigate the playlist track-by-track or randomly. With voice navigation, users can now make the Shuffle easily usable in more contexts (at the gym, for relaxation, language study). The ability to navigate without looking at a display is an important new feature because it integrates well with activities like driving where a lot of displays are already fighting for the driver’s attention (speedometre, GPS controls, etc..), and I’d expect VoiceOver to be rolled out across the entire iPod and iPhone line at some point. The end game for portable devices is full voice recognition, but can’t really be done economically at present (smartphones are looking to do voice recognition by sending the voice command over the Internet to a powerful server which deciphers it in real-time). In the meantime, a simple 3-button control plus VoiceOver might be a decent compromise looking ahead over the next few years.

In summary, Apple does appear that it can indeed roll out evolutionary products in its product line, with recent announcement about new computers (iMac, Mac Pro) and now new low-end iPods. The question mark is how Apple manages the roll-out of a brand new product segment, and it looks like that point will be coming up later this year. Reuters is reporting that Apple has ordered 10″ touchscreens for the Q3 timeframe, which does suggest that a tablet netbook is coming from Apple. If Apple can deliver a killer product into the high end of the netbook market (maybe in the $500-800 range), it could help Apple to fill a gap at a time when consumers are pretty price-conscious due to the economy. The important thing for the stock (now trading at 17.2 forward earnings) is that this should result in some positive news flow in anticipation of the new tablet… Whether this new tablet will be any good is another matter for debate, but the key is that we won’t know for at least a few more months until some beta hardware gets leaked, which should at least generate positive news flow for Apple and for the stock.

Leave a Comment

Filed under News

Tags: , ,

March 5, 2009

Clearwire’s prospects and why WiMAX might be the Blu-ray of the wireless world

Clearwire is expected to reveal its deployment plans today after the market close, and it’s a good time to look at Clearwire’s prospects in the market. The news of Clearwire investors getting additional shares as a result of the stock dipping below $3 hasn’t been good for the stock, and it’s down more than 11% on the day. This despite ABI Research’s expectation that the market with grow by… 4500% in 2009 (I hate growth numbers off a very small number). “To ignore a growth market in a down economy would be a mistake” notes ABI, but the real question is whether WiMAX is going to do better or worse than expectations, and device availability seems to be a critical consideration in thinking about future success of a new technology.

Where does WiMAX fit into an emerging mobile broadband market? The comparison in the title of this piece to the Blu-ray technology seems to be quite appropriate in my view – WiMAX offers high-speed mobile connectivity to devices, and is an improvement over existing Wi-Fi because of better mobility due to improved coverage (in theory) and faster speeds than existing access over the cellular network. The openness of WiMAX was a big bonus for Google, as providing a la carte wireless broadband service to consumers is closer to how people buy fixed broadband access. At the end of the day, however, does it matter that much to consumers? Wi-Fi was a huge success years ago because it gave users the capability to access the Internet without being plugged in, but WiMAX as an incremental service won’t give a huge benefit for most users over conventional Wi-Fi (at cafes and in buildings) or cellular broadband (through incumbents). In a sense, this mirrors the advantage that DVDs had over VHS Tapes (much clearer digital picture and quality didn’t degrade with repeated viewings)… for a lot of users, Blu-ray HD discs don’t offer a quantum leap in terms of a better viewing experiencing, which explains why consumers aren’t burning their existing DVD collections in the streets.

Alcatel-Lucent’s presentation at CeBIT in Germany hits the nail on the head in describing WiMAX’s role in the emerging market: “We see it more as an extension of the DSL (Digital Subscriber Line) footprint. We don’t think it’s a viable option to compete against LTE as a mobile only activity, but other people may have different views.” WiMAX as an improved DSL to connect the unconnected makes a lot of sense, especially in areas like India and Africa where the infrastructure simply doesn’t exist today (and with sufficient density to make the service worthwhile). Alcatel-Lucent has already signaled a move to lower its investment in WiMAX in order to focus on LTE (cellular-based), and this change in focus makes sense. As was the case in DSL, not a lot of companies are needed to serve the market and does WiMAX need to have a dozen companies offering similar products? With Nortel now out of the market, Harris Stratex buying Telsima for $12M, Cisco buying Navini, Redline going through a significant downsizing, the market is shrinking and consolidating.

The biggest problem I have with Clearwire’s business model thus far is in targeting major metropolitan areas (Sprint’s Xohm which has been merged into Clearwire, launched trial services in Chicago and Washington, and it’s expected that Clearwire will be targeting major cities for the next wave of its rollouts. The challenge for Clearwire is that these cities are already well-covered with fixed line fiber access lines, cable Internet, and cellular access. A better target might be rural areas, with choices in those areas being dial-up service and satellite services like WildBlue. But making the services profitable in these areas is tough given limited population density. Is Clearwire the answer to a question that hasn’t been asked yet? Perhaps. But that’s not to say that a completely new offering can’t be successful (iPhone was a game changer in the smartphone market). The problem might be that Clearwire will have a tough time getting the service into users hands.

The notion of “getting devices for free” isn’t working out so well for WiMAX. Dell and HP, the two leading PC makers in the U.S. with a combined market share north of 50%, aren’t readily supporting WiMAX. A search on the HP site didn’t find any options for WiMAX, while a press release did note that Dell was making WiMAX available on several laptops but again I did not see find a single model with the option to buy WiMAX available. While Dell was selling regular integrated Wi-Fi with its laptops, HP did have the option to choose a Verizon cellular data card with its laptops – given that customers can at least experience nationwide coverage on Verizon’s CDMA network, HP does realize that today, selling data access through Verizon makes sense for its customers… today. Moreover, HP is only selling the data cards with a data plan, which is certainly providing HP with revenue on the service side as well – In contrast, Clearwire’s value proposition in the U.S. is essentially pay-as-you-go, and can’t really afford to pay PC vendors like HP an equivalent amount.

I finally did find WiMAX offered… at Lenovo (which includes the former IBM ThinkPad division) for $80 which includes Wi-Fi capabilities as well, though Lenovo recommends getting the $40 option for just 802.11a/g/n. In addition, the WiMAX option is simply included in the “Integrated WiFi wireless LAN adapters” section, which must be confusing for customers – “What is WiMAX and why do I need another WiFi interface?”. The biggest problem with WiMAX is perfectly illustrated by Lenovo’s ordering section for “Mobile Broadband” options. The AT&T Wireless broadband option is included by default! While the option actually is $30 (the customer would get $30 back if they choose to not select AT&T) and Verizon’s data access is effectively $150, this is a significant detail to mention. AT&T is throwing in a month of free access in return for the $30, providing a solid incentive to try the service is they don’t have broadband access already.

lenovo1

The conclusion, at least in the U.S., is that the WiMAX consortium isn’t doing a great job of laying the foundation for building WiMAX demand. When Intel had its big success with the first Centrino platform years ago which included Wi-Fi, the lure of a low cost upgrade to get Wi-Fi was a no-brainer for consumers – this made standalone Wi-Fi cards obsolete for the most part. The hope back in the 2005-2006 timeframe was that WiMAX would help to solve the chicken-and-egg dilemma that new technologies face, by giving away WiMax for free. Well, an $80 on a laptop means that users won’t buy Intel’s integrated WiMAX unless they need to use it today. Intel might be guilty is of not doing what’s necessary to get WiMAX really going because it could be giving away this capability for free, though for a company looking for margin expansion, it’s been understandably reluctant to do this. As discussed above, the availability of cellular-based wireless broadband with lock-ins and incentives for laptop vendors might even be a bigger problem – the incumbents appear to be winning this war by throwing $$$ at the big PC makers, at least so far.

Leave a Comment

Filed under News

Tags: , , , , , ,

March 3, 2009

A pricing war is emerging amongst mobile operators – A good thing for smartphone makers

Just over a year ago, the mobile carrier world was in flux, with all of the major carriers launching $99 unlimited voice plans. This was an important moment in time, and signalled that a paradigm shift was coming – from landline phone calling to wireless mobile phones. As expected, consumers have abondoned their landlines, especially given the macro economic climate we’re in now.

When earlier in the year, Sprint’s pre-paid division Boost Mobile started offering $50 unlimited voice plans with nationwide coverage, the fear was that it would trigger a price war amongst the Tier 1 carriers (Sub $50 plans were already being offered by smaller carriers like MetroPCS and Cricket but with limited geographic coverage). A price war is now almost certainly around the corner.

T-Mobile is now offering $50 unlimited voice plans for loyal customers (22 months of previous service is needed to qualify) and can combine this with unlimited data for an additional $35/month. An $85 unlimited voice + data plan might be a new benchmark for the Tier 1 vendors. To compare, I went onto Verizon’s site, and unlimited voice + data looks to be around $130/month. AT&T is around the same price, and unlimited voice and Apple iPhone or BlackBerry service is $130/month, and Sprint’s Simply Everything plan is $99/month. T-Mobile’s new plan ends up being 15-35% cheaper than comparable plans (if you don’t factor in handset availability – iPhone is still only available on AT&T in the U.S.).

With Sprint and T-Mobile now offering substantively cheaper plans than AT&T and Verizon Wireless, the big 2 may now need to address the situation, and may need to bring down the $130 price ceiling closer to $100, which is a mixed bag for the wireless carriers. While this could accelerate the landline-to-mobile shift (bringing more gross additions), it now brings unlimited usage plans to a greater percentage of the subscriber base. For owners of base iPhone plans in the $80 range, going to $85 or $100 isn’t much of a stretch to enjoy unlimited nationwide calling – further downwards pricing pressure on all voice + data plans is a reasonable expectation. For operators with a lot of exposure to landlines like AT&T, this negative trend has been around for years and might be accelerating to some degree.

So, while further irrational pricing in the mobile phone area makes for more risk and an overhang for valuations amongst carriers, this trend is positive for smartphone makers (which are up more than 5% today), especially as an NPD report is showing that smartphone share has grown from 23% from 12% last year, and 3G phones have grown their share from 46% last year to 66%. In short, with customers enjoying smartphones features like faster connectivity and touchscreens, and cheaper voice and data plans, the smartphone is increasingy becoming the consumer’s main device for communications, entertainment, and productivity. And looking at the players within the smartphone market, I think the number of devices is going to rationalize over time, just as it did with the mainstream phone market. Looking ahead, I believe Apple’s iPhone and RIM’s BlackBerry will remain premium devices in the market, and the question mark for me is which other devices will make it this year in terms of market share. Nokia has had some trouble in merging its numerous acquisition into a coherent new platform, Microsoft simply isn’t on enough hot phones to make a difference, and Google’s Android available on T-Mobile as the HTC G1 has promise (and is rumoured to be on a large number of new phones this year). The wild card is Palm’s Pre, which has received great reviews and has enabled the stock to quadruple off of its bottom, but isn’t even in the market yet. At the end of the day, this market will likely shake out over the coming 12 months with maybe 3 or 4 survivors…

Leave a Comment

Filed under News

Tags: , , , , , , , , , ,

February 26, 2009

The rationale for Nokia’s move into the notebook space and is Apple certain to dump Intel as it rolls out a tablet/netbook?

In an interview with Finnish broadcaster YLE yesterday, Nokia’s CEO Olli-Pekka Kallasvuo signalled that the company was looking to get into the laptop PC market, noting that the lines between the PC and the smartphone was starting to blur – “We don’t have to look even for five years from now to see that what we know as a cellphone and what we know as a PC are in many ways converging.” No doubt, Nokia must be feeling left out of the party as Asus has already indicated that it is exploring putting Google’s mobile operating system Android on its line of netbooks, while Apple has a common operating system across both PCs and smartphones (OSX) and Microsoft having “related” products in Windows and Windows Mobile. With Nokia open sourcing its mobile OS (Symbian), how does Nokia improve the chances that Symbian gets some airtime in both the smartphone arena and in an emerging tablet netbook market? By bringing out a PC product of its own.

The pressure on Nokia to bring out such a product may also be coming from its carrier partners, which are increasingly giving away netbooks from companies like Asus and Acer for free when customers purchase a mobile broadband subscription. For service providers like Clearwire in the U.S., they are simply selling broadband on the go, which is only really enabled in high-end smartphones – a device that they can’t just give away for free. Put another way, Nokia has been missing out on an opportunity to keep its brand front and centre with millions of new customers per quarter… a situation it needs to rectify as soon as possible.

What would a new Nokia netbook look like? The first question is whether it will deploy an Intel chip or not. On one hand, Nokia might not like that Intel is working with Option and Ericsson to integrated HSPA in future chipsets, and could instead choose to work more closely with its new friend Qualcomm to build an ARM chip along with a Gobi chipset to enable access to mobile broadband networks. Symbian will work well on ARM chips, and this would open up access to its emerging Ovi ecosystem. The extra horsepower and larger screen available in the netbook/tablet would also make web browsing and basic productivity apps a good experience. Microsoft, which can already support ARM chips on Windows CE, could be pushed to make Windows 7 functional on ARM if enough manufacturers threaten to go to Linux… as they did with the first round of netbooks. Could Nokia decide to just go the way of Intel Atom + Windows 7? Absolutely. This is an option, and Nokia could release an device using Intel Atom along with a Gobi chipset from Qualcomm (which can support both GSM and CDMA flavours of 3G), though battery life is an issue, and the most recent Oqo 2+ only gets on the order of 3 hrs of battery life if stretched to the maximum. At the end of the day, entering into the PC market is risky especially because the market demand is already satisfied with existing products. While going the way of a Taiwanese ODM might get Nokia to market slightly faster, just producing another low-end PC doesn’t seem to be worth the trouble especially from a profit-perspective, which suggests that Nokia might be looking at controlling more of the design and perhaps going to an ARM-based processor for its tablet netbook.

What impact would this move have on Nokia’s financials? I’ve done a bit of scenario analysis based off of Nokia’s most recent Q4/08 results, and if Nokia is able to sell around a million devices a quarter (out of a total market of 3.6M devices/quarter), the impact on Nokia’s corporate gross margin (and operating margin) could be on the order of 20-30 basis points… perhaps not too big an impact given that it’s a bit of a land grab out there right now. Keeping the Nokia brand and ecosystem in front of customers will be critical as this theme plays out.

nokia2

In summary, given the moves by Google, Microsoft, and Apple as smartphones increase in size, Nokia entering the small notebook market isn’t too much of a surprise, especially as carriers look to take their piece of this market. For Intel, their lack of preparedness in securing their positions in both markets is somewhat disappointing, and punting their PXA application processor unit to Marvell in 2006 doesn’t look so good in hindsight.

And as I dig deeper into this issue, a thought on Apple’s positioning in this area. Is Apple sure to move off of Intel chips as it looks at a tablet netbook? I’m not so sure anymore. On one hand, going to an ARM processor (built by PA Semi or by Samsung) would be good for Apple’s margins and could help improve the performance capabilities of the device from a power perspective. Apple certainly has the capabilities to switch architectures across the board, though its “Universal Binaries” approach, which enables progrmmers to build in support for a new processor family by just recompiling the code (UB’s currently support the PowerPC family of CPUs from IBM and Intel CPUs). This could mean that Apple could replace its processors across the board – from the low-end to the high-end. However, this would take time, and Apple would REALLY need to get into the chip business in that case. I don’t think consumers and enterprises would be open to the incompatibilites that this would entail… but it’s at least possible over time.

What Apple really wants is a bargaining chip with Intel, and the power to get custom chips from Intel that don’t get sold to other companies (a low profile chip was made for the MacBook Air). With Apple also controlling the programming environment for Macs and iPhones, Apple has the flexibility to seamlessly port iPhone apps to a Mac platform or vice-versa. Perhaps we’ll see Intel’s new Atom replacement CPU Pineview (available in Q3/09) on an upcoming Apple device, and performs a small form-factor one exclusive to Apple to boot. For Intel, it’s dominant share of the existing PC market does have value as the notebook market evolves into smaller form factors along with touch screens, HD video, and ubiquitous mobility, but the pressure will be on to deliver in 2009 – one thing seems certain… 2009 is the year that we see mobile processors from ARM make their way into the ultra mobile PC and MID (Mobile Internet Device) market, much to the chagrin of Intel.

Leave a Comment

Filed under News

Tags: , , , , , , , , ,

Older Entries