David M Smith

A member of the Gartner Blog Network

David Mitchell Smith
VP & Gartner Fellow
16 years at Gartner
30 years IT industry

David Mitchell Smith is a vice president and Gartner Fellow in Gartner Research, where he specializes in the impact of catalytic technologies such as the Internet, Web 2.0, cloud computing and consumer technologies. Read Full Bio

Nokia dilemma about geography and culture

by David M. Smith  |  February 10, 2011  |  Comments Off

Everybody’s weighing in on the big Nokia news expected tomorrow.  This isn’t about the actual choice widely expected to be be that Nokia and Microsoft reach a deal around Windows Phone 7. I actually predicted this when Stephen Elop joined Nokia. So that’s not surprising to me at least. Although the speed at which the change (less than 6 months after joining) is a bit of a surprise. But obviously time is of the essence here.

What has been fascinating is the divide between North American and European perspectives. The reactions to the leaked “Burning Platform” memo by Elop.  Denial and claims that the memo isn’t real and must have been written by an American analyst.  Well, it was written by a (North) American (Elop is actually Canadian).  And it looks real to me.

The point is that geography and culture matter a lot. What I am writing is absolutely influenced by the fact that I am an American.  But an American who has seen technology trends worldwide.  As an American (hopefully not ugly), I survived years of phone envy.  For the longest time, Europe with its GSM standard was years ahead of North America in usage and advanced usage.  I’d go to Europe and want the phones – in fact I had a Bosch worldphone on Omnipoint (the precursor to Voicestream, the precursor to T-Mobile) so I saw what was happening.  But I’m also a web guy.  And despite the advances, the mobile web (e.g. WAP) was still awful even in Europe.  Every year I would go over and expect to see something great because year after year, people were talking about the ‘year of the mobile web’. Every year I was disappointed.  The latest and greatest barely worked.  A shadow of the desktop web. Similar story in Japan.

Well, it finally happened.  In 2007.  With the iphone. The first real mobile web experience. And than apps happened. and then the tides turned and the most advanced use of mobile phones and the new developments were happening in North America, not Europe.  The vendors gaining credibility, share, and power were North American vendors and the North American market (which Nokia had lost even before the iphone) became much more important. So for Nokia, getting a presence in North America (in addition to the technology and UI and other issues) became paramount.  Enter Elop.

So geography matters. and attitude as a result of that geography matters.  Being on top for years leads to complacency, denial and and myopia.  There are obviously lessons to be learned here far beyond just Nokia, smartphones and even the tech industry.

Comments Off

Category: Uncategorized     Tags:

Windows Phone is “Too Big to Fail”

by David M. Smith  |  November 11, 2010  |  Comments Off

When Steve Ballmer was interviewed at Gartner Symposium in Orlando last month, he said that his riskiest product bet is the next version of Windows.  I agree with him.

What’s not obvious is what is on the critical path to that product.  I believe it is Windows Phone 7. Here’s why: It’s simple. it’s the Ipad and how Microsoft has to compete with it.  Taken to extremes, without a competitive response, the Ipad and other tablets will have a huge negative effect on PCs as we know them today.  Thus far Microsoft’s response has been very quiet, and with a little talk about some interim, enterprise focused Windows 7 tablets that derive from the company’s earlier tablet efforts which were of very limited success.

I believe they are quiet about the response for a number of reasons.  First, they are focusing on consumer markets first, where Apple has set the bar for secrecy awfully high.  They can’t get there but they are trying. someday I’ll write about that more…  Also, they are simply waiting to see how Windows Phone 7 and the touch enabled apps for it fare in the market.  Success with Windows Phone would mean that they can move forward with an offering that leverages those apps.  This is the strategy employed by Apple and being employed by Google and HP/Palm.

The next version of Windows, although I don’t know that much about it, I’m expecting to be some kind of melding together of the Phone and the PC at the OS level.  What kind of programming models exposed are more important than an underlying merged kernel and critically  important to any platform strategy.  Indications have been that a web-based Javascript oriented HTML5 heavy strategy is a major part of that strategy (similar to Palm’s WebOS).  However, to be competitive in tablets, lots of touch-enabled apps are required.  Eventually, HTML5 apps over the next few years will be more important, but for now that means primarily Silverlight apps that can be leveraged from the Phone.

So there is a lot at stake with Windows Phone 7.  It’s not just another phone.  Even though the early indications are that these things aren’t flying off the shelves, don’t expect Microsoft to give up.  This is not “Kin 2.0”.  They have bet and bet big and will invest in engineering, marketing and whatever they need to do to make it successful.  They can afford to do it.  More importantly, they can’t afford not to.

Comments Off

Category: Uncategorized     Tags:

Microsoft PDC10 shows big changes

by David M. Smith  |  October 29, 2010  |  1 Comment

As a platform company at its heart, PDC is a great way to understand what Microsoft’s real strategies are.  In addition to its focus on three trends that are popular overall (HTML5, mobile, and cloud), the specifics show some significant changes and enhancements to the the company’s strategy.

First, HTML5.   Microsoft led off with this as the lead message for developers.  If there was ever a huge change, it is here.  Only a year ago Microsoft was very tentative about HTML5.  Now it leads off PDC with it.  Perhaps a bit overstated or futuristic, it showed charts espousing HTML5 across multiple devices including mobile and Xbox. Interesting in that it has not yet formally announced intent to support HTML5 on these devices yet.  We’d already seen that Windows and IE9 have moved towards HTML5 in a big way. However, messages here at PDC signal that the company’s cross platform strategy is now HTML5, not Silverlight.  Microsoft appears to have read the handwriting on the wall and seen that HTML5 is the future and that fighting it with Silverlight on the web is a losing battle and has therefore cut its losses.

Silverlight may be downplayed as a brand, but it remains a very strategic technology, forming the basis of the Windows Phone 7 platform and I think will likely play a role  in Microsoft’s future desktop and tablet offerings. As for tablet, I believe Microsoft is delaying talking about its tablet strategy until it is clear if Windows Phone 7 is successful or not.  If successful, I expect to see its strategy look and feel a lot like that (as most tablet strategies are leveraging touch technologies and apps across both).  If it doesn’t succeed, well, lets just say that it wouldn’t be very good news for Microsoft in much more than the mobile phone space….

Second, mobile.  Not that much totally new here that hasn’t been shared in detail over the past few months.  However, PDC provided a good showcase for the new Windows Phone 7 platform and its free and excellent tools as access to data sources through odata.  They showed  some of the over 1000 apps  available and announced an exclusive feature (Amazon Kindle store). Giving away phones to the developers was very popular and will help drive more apps.  There was still no formal announcement of HTML5 support for mobile, but given the messages on HTML5, it is just a matter of time.  Microsoft appears to be holding off as it is encouraging the native Silverlight development for now as its app catalog is just starting out.

Third, cloud.  In Microsoft’s case that is of course Windows Azure and related pieces.  For the first time, Microsoft has aggressively positioned Azure as PaaS.  It is a legitimate use of the term, and as is often the case with Microsoft offerings, it doesn’t fit neatly into categories.  Azure has many IaaS features and is primarily an OS.  But OSes usually have platform capabilities.  And Microsoft as a platform company is not surprisingly positioning it this way.  What’s surprising is that it has not done so as strongly until recently.

Azure has had some success thus far, primarily with ISVs and small business. Not so much with enterprises.  This will change as a result of some of the announcements at PDC.  What they have needed are ‘onramps’ to the Azure cloud.  In Microsoft’s case, their private cloud offerings do not provide this very well.  Its Windows Azure Platform Appliance is not yet available and it is not clear who will be able to use this as a private cloud.  Its System Center driven Windows Server virtualization offerings, while useful, don’t really provide much of a path to public cloud.

Two major features announced at PDC change this and provide those onramps.  First, the long rumored best-kept secret of the VM role.  This will enable existing Windows apps to run on Azure without reworking, a welcome feature for enterprises not ready to rewrite apps for the cloud.  Second, is a multitenant container to be launched by the end of the year.  Called the Appfabric container and composition model, it is significant for two reasons.  First, as an onramp.  Appfabric is meant as a way to write programs that span Windows Server and WIndows Azure.  This is an onramp to the cloud for new apps.  The other significance is that Microsoft is jumping into what some might call ‘real’ PaaS – a multitenant shared everything model. This in many ways will legitimize this approach as a major enterprise credible company has endorsed this approach, in contrast to Oracle’s recent snubbing of the concept. 

So, big changes in web and cross platform strategy towards HTML5.  Showcase of Windows Phone and the future of Silverlight as a native technology beyond phones. And some very welcome enhancements to its Azure cloud that will begin to solidify the company’s position as an emerging leader in cloud and drive enterprise interest.  All in all, a pretty significant PDC.

1 Comment »

Category: Uncategorized     Tags:

Twitter as the CB Radio of today

by David M. Smith  |  September 24, 2010  |  1 Comment


I’m not saying that Twitter is a fad (as some said of the Internet way back when).  Just trying to make some connections. Two recent events (OK, not so recent as they were a few months ago but hey, I’ve been busy and finally decided to do some more blogging…) made me think about twitter and CB radio. Those of you old enough to remember the CB radio craze of the 70s (if not, Google it…) will get it. The two things are:

1. an announcement of Twitter for cars, which is where CB radios were  used (actually started in trucks, where they are still used)..
2. Tibco’s introduction of Tibbr, an enterprise focused twitter like offering (somewhat like yammer, but with a different usage model).

What’s the connection? Its about how these things are/were used.  I’m pointing out some of the similarities regarding what they are/were actually USED for.  Twitter is used for many things including to broadcast lots of information on a public media (the Internet) that can be "tuned into"  various channels, kind of like CB.  Certain channels were used for certain things, by convention, with little formal control. Similarly, Twitter has #hashtags which provide a type of convention as well.  .

Tibbr is somewhat similar to twitter but is designed for internal enterprise use and as such is more focused on following subjects, not just people. This approach may indeed have more direct relevance in a business context (think Web 2.0 style tagging, next gen knowledge management, etc. ).

While many use Twitter to "follow" people, I find that I use it much more to tune in on topics.  Like trending topics and using it to monitor what is being said during an event.  This includes
Gartner events, announcements, trade shows, etc.  Actually I find the most compelling use is to tune into what people are saying during sporting events – it seems to amplify addictive behavior (tweeting and watching sports).  I enjoy watching the Red Sox (even this year) and tuning into the #redsox hashtag on Twitter. I can find out not only what the color commentators are saying, but what they’re not saying and what people are saying about the color commentators. and the commercials.  It really enhances the experience.  Now that we have the emergence of TVs with Internet connectivity and widgets I expect to see this type of activity increase.

What are you using Twitter for? Like CB?

1 Comment »

Category: Uncategorized     Tags:

Oracle spins its cloud strategy 360 degrees

by David M. Smith  |  September 20, 2010  |  Comments Off

Oracle Open World opened yesterday with a keynote that included CEO Larry Ellison talking about its cloud strategy.  It wasn’t very long ago that Larry was quoted saying some not so nice things about the term cloud. So what’s happened? Has he changed his mind?  Details are sparse and will likely be expanded upon later in the week so this post will be fairly sparse as well.

Some have interpreted this as Oracle getting into the cloud.  But is this a 180 degree change of strategy?  No, not really. So far it looks more like a 360 to me.  The full spin, so to speak, all the way around back to where it was before.

spin cloud

Not that Oracle isn’t building some good stuff or that people will not find it useful.  The point is that other than some continuation on the same path (enabling, not providing cloud services, focus on private over public cloud)), the strategy is the same. 

Before yesterday, if you ask Oracle what their cloud strategy was the answer would be essentially ‘you can build clouds with our stuff’.  Now it’s still the same.  But more stuff.

But now it’s cloud. Poof.

Comments Off

Category: Uncategorized     Tags:

Google and net neutrality: Head Fake or “If you can’t beat ‘em join ‘em”?

by David M. Smith  |  September 15, 2010  |  Comments Off

There has been much talk of Google by way of its partnership with Verizon vis a vis Net Neutrality.  Many opinions describe it as Google going against its long standing “don’t be evil’” mantra. No comment here on that topic.

However, Google has indeed backed the concept of net neutrality for a long time.  It is not for altruistic reasons, but rather because the more people use the Web, the more they view Google ads, which is where 97% of the company’s revenue comes from. It has viewed the status quo of a very open Internet as a good thing for its core business and it drives much of the company’s strategies.

To that end, Google has employed tactics that are best described as “Head fakes”.  These have included bidding for wireless spectrum with the goal not to actually own it but to open up the terms and conditions. With Android, it launched the Nexus One phone with the goal to erode wireless carrier influence, not to actually be a manufacturer of wireless devices.  Ironically, with the failure of the Nexus One strategy and the overall thus far success of Android, the influence of wireless carriers has increased.  Not exactly what they had intended.

So now, Google and Verizon.  What’s really going on here? Yes, Google is close to Verizon as a result Android.  And that is likely at least a factor.  However, the re is more than meets the eye here as US regulators don’t seem to be interested in decisively acting on net neutrality.  So is Google doing another ‘Head Fake” by attempting to drive them into acting by partnering with Verizon?  Or have they resigned themselves to the reality of carrier power and figured that “if you can’t beat ‘em, join ;’em”?

Either way, works out well for Google.  Doesn’t make them evil. Maybe not quite brilliant. but definitely pragmatic.

Comments Off

Category: Uncategorized     Tags:

Elop to Nokia. Will Nokia embrace Windows Phone 7?

by David M. Smith  |  September 10, 2010  |  3 Comments

News broke overnight about Microsoft Office executive Stephen Elop moving to Nokia as CEO. This by itself is shocking news, but what other shocking aftershocks might follow?

I’m no Nokia expert. All I know is that my first mobile phone was from Nokia back in the 90s but that nothing they’ve made recently has been compelling enough for me to seriously consider it.  Many have given up on Nokia, especially in the US where most of the smartphone action has been.

Microsoft, I’m much more familiar with.  Had one of the very first Windows Mobile phones before they even called them by that name. For a while, they were competitive, but as time went on, others, notably Apple and Google left them in the dust.  Many have given up on Microsoft in mobile. But Windows Phone 7 is about to launch. And I’m not ready to write it off.

The market has spoken and said it likes the idea of alternative, more open offerings than those from Apple.  It also still likes Apple, but a lot of the Android success is due to offering more choices than Apple does (carriers, form factors, etc.).  But with that openness comes tradeoffs.  Ironically closedness due to increased carrier power.  A bit of fragmentation.  Things that Microsoft has learned the hard way through a decade of Windows Mobile mistakes.  Google is learning them now.  So, timing for Microsoft with a hopefully ‘just right’ balance between control and open, could be very good.

Elop, in his short time at Microsoft, was involved in deals that put Office and Silverlight on Nokia phones.  There’s been a lot of ‘advice’ given to Nokia that has suggested they drop everything and go Android.  Here’s my radical thought: Nokia will go Windows Phone 7.

Many have written each off.  However, together they have a much better chance.


Category: Uncategorized     Tags:

Is PaaS Passe yet?

by David M. Smith  |  September 2, 2010  |  2 Comments


Disclaimer: I’m ranting about the term PaaS, not the concept.

We just published the 2010 Hype Cycle for Cloud Computing and is it a surprise that the most hyped term (other than cloud computing itself), is PaaS? This is in part due to legitimate excitement about some of the offerings and possibilities.  However there is a huge amount of confusion generated by this term.  There are many problems with this term and the absurd discussions that come about keep getting more absurd all the time.

I can’t even count how many conversations I’ve had over the past year or so explaining why hosting something like Websphere or Oracle on Amazon is not PaaS. (A good rule of thumb is if you need a traditional software license it’s not a service.) You have some software being hosted on some cloud infrastructure service (IaaS).

I’ve been meaning to write this for a while. Many of you who talk to me have probably heard it a few times already. A couple of recent discussions have prompted me to finally do it.  First, Red Hat’s “PaaS” announcement which is not a PaaS but enabling technology that can be used to build one.  But what really put me over the edge was a conversation about another vendor’s offering of "PaaS as a Product". To me this is as absurd as General Motors, who builds cars that can be used as taxis (and could be called ‘car as a service’ (or CaaS)) referring to
their cars as "CaaS as a car" or "CaaS as a product".  Pretty absurd isn’t it?  That’s just the beginning of the absurdity that the PaaS term leads to.

It comes down to use of two words, both of which by themselves are often sources of confusion (and in the cloud world, we thrive on confusion, but this is too much). The words are Platform and Service, which of course are the P and the S.

First, lets look at “service”.  Most of the offerings from companies like Redhat and VMware are cloud enabling products, not services. Not that there’s anything wrong with that…We even have a term for it at Gartner – Cloud Enabled Application Platforms (CEAP).  They can be used by service providers and even internal data centers to provide a "PaaS".  But they themselves are not services.  (Side comment, I’ve yet to see true "PaaS" or SaaS internally but that’s a rant for the future…)

Next lets look at “platform”.  Nobody could say it better than one of my favorites, Dilbert: .dilbert
It’s true that everything can be a platform, and that it can be an incredibly confusing term, especially when used unqualified like in PaaS. However it can also be a very helpful term and concept when viewed as a relative term.    The example I often give is platform as the "you are here" marker in IT architecture. Like when you enter a city you are unfamiliar with, you look for a map with some indicator of where you are. Without it you have little hope of getting to your destination.  Similarly in IT, the term platform can be used that way to help clarify, not confuse.  As there can be many kinds of platforms (eg, hardware, browser, OS, database, app server,…), qualifying it is imperative to use if to help, not confuse.

Platform refers to an extensible entity, something that is built upon. It resides on top of infrastructure and below applications (also relative terms). At Gartner, we have referred to all of the levels (not just "PaaS") as types of web/cloud platforms.  After all, don’t many of us also refer to Amazon as an IaaS platform when we are talking about building on top of it? What if what we
are building a "PaaS" on top of it? What would we call that? a PaaS Platform? Here we go getting absurd again.

Best to my recollection, the term PaaS can be credited to (blamed on?) Salesforce.com. Their entry, force.com, is indeed delivered as a service so it’s really not their fault. However, as cloud terminology evolved and we had IaaS and SaaS, using PaaS as a general term seems to be what happened. And NIST adopted it and everybody uses the term.  Even us, although I do it kicking and screaming.

A better term to describe what most people mean when they say PaaS is "middleware in the cloud".  This doesn’t use the term platform as ambiguously and it makes no claims that anything is delivered as a service at that level. And when you want to actually mean services, we call it "Cloud Application Infrastructure services"

I’m not betting that the term PaaS will disappear (Certainly the concepts won’t as they are critical.) It’s too entrenched.  But in the confusing world of cloud computing, is the term helping or hurting? I say it’s hurting.


Category: Uncategorized     Tags:

Google me? You’re kidding me…

by David M. Smith  |  June 30, 2010  |  Comments Off


I am one again amazed at the buzz (sic) generated by the prospect of Google once again entering the social networking space.  Originally the rumor was attributed to Kevin Rose of Digg (the link appears to be dead now, but it was picked up other places like here.

But, isn’t Google a ‘web 2.0’ poster child company?  If so, how can it continue to have to enter the social networking space (which means they haven’t yet succeeded), yet the conventional wisdom is that they are a successful web 2.0 company?  This same conventional wisdom equates web 2.0 with its most visible component, social networking.

A similar thing happens in cloud computing.  Once again, Google is often considered a cloud computing poster child.  However, if you also look at the way that the same conventional wisdom views the cloud as these limited buckets of IaaS, PaaS and SaaS, then Google ends up not being a player in any of those spaces, if you look at revenue (97+% of Google’s revenue is derived from search and advertising, which is not in those –aaS categories).

So, how can Google be considered leaders in web 2.0 and cloud computing given these oversimplifications? Just askin’

Comments Off

Category: Uncategorized     Tags:

Big implications for the web and the cloud at Google I/O

by David M. Smith  |  May 19, 2010  |  1 Comment

I spent today at the Google I/O developer conference. Fun Fact: Google claims it stands for Innovation and Open.  Day one (unfortunately only able to attend one day) had two main themes: one was about the web, the other about the cloud. Big implications for both.

The Web, including HTML5, video and a web app store:

  • Lots of demos of HTML5 web app capabilities including Darkroom, clicker.com (and their new 10 foot UI clicker.tv) and a really slick sports magazine by Sports Illustrated that looks like it was made for an Ipad.  Drag and drop into Gmail, and of course lots of video. The goal was to demonstrate how rich HTML5 apps can be.  Very rich indeed. 
  • WebM as a new web media initiative (my assumption about the M) was announced and its main component is the VP8 codec. As was widely speculated, Google did make VP8 available as open source.  As there are issues regarding IP regarding other codecs, VP8 is a good fit for HTML5 video.  Mozilla and Opera were on stage to support it. Not coincidentally, Microsoft announced support for it today as well.  We’ll see what Apple will do.
  • Adobe was also on stage showing new features of Dreamweaver that make developing HTML5 easier especially on different form factor screens.  In addition, they announced that VP8 will be built into the Flash player and hinted that we’ll see more Flash stuff tomorrow.  I have been expecting to see Google and Adobe get much closer so tomorrow could be very interesting.  Flash represents a way for Google to transition from native mobile apps (eg Android) to the eventual pure web that they envision.
  • The Chrome web store was also announced.  This is the sleeper announcement.  Missing from the web has been a way for people to find apps and for developers to distribute them and monetize them.  Google is looking beyond advertising monetization models and for now is focusing this on html5 web apps for Chrome and ChromeOs, free and paid apps.  Today it is not focused on mobile, but IMHO just a matter of time.  And will be one of the keys to adoption and acceptance of mobile web apps as a viable option, not just native web apps.

The other main theme was “Bringing the cloud to the enterprise”. 

  • First, VMware CEO Paul Maritz talked about Cloud portability and the role that Java, specifically the Spring framework, can play here.  Through technologies that VMware acquired from Springsource, it has begun to drive Spring as a viable way to develop Java apps for the cloud.  It spans not only the new version of Google App Engine (for business) but also its recently announced VMForce joint venture with Salesforce.com, as well as private clouds running Vsphere and other public clouds running Vcloud. VMware is posed to fill the void in enterprise cloud leadership.  While there are certainly cloud leaders in general (e.g., Google, Amazon), there are no natural enterprise leaders.  Microsoft is trying, but it is early. 
  • GAE for business was also announced as an enterprise focused offering that takes GAE and adds the spring framework support along with management tools, SLAs, SSL, a SQL database, and revised pricing.  The pricing model derives more from Google Apps (SaaS-style subscription vs. IaaS-style ‘pay as you go’ which is how GAE (non-business) is priced.  It is interesting that this pricing model was talked about, whereas for VMForce, it was not finalized.  I view this as an indicator that the SaaS subscription model will be more popular for “PaaS” style offerings such as GAE.

Clearly big news on both fronts.  More to come.  Oh and more tomorrow but I’ll have to get it second hand.

1 Comment »

Category: Uncategorized     Tags: