Here’s what I think its taking place:
Microsoft executives and Board members are no dummies: they know Cloud Computing threatens the Windows + Office + Exchange gold mine, the biggest in our industry’s history. They know the future is Office + Exchange running in dual-mode. From the Cloud when a Net connection is available; locally when the Cloud is out of reach. Everything synched back when the connection is restored. Imagine Outlook in Cache Mode, just with a browser, without a local client, generalized to all Office applications. Their delicate mission, should they choose to accept it, is to move Office and Exchange into the Cloud, into dual-mode applications. The challenge is to get there before Google Apps gain acceptance but without prematurely cannibalizing the existing Office + Exchange profit stream.
On its side, Google wants to protect the search-based advertising gold mine. To do so, they need to hurt Microsoft’s ability to finance a broad-front attack against Google’s core business. That’s why Google wants to offer an alternative to “Office in the Cloud”: with Microsoft no longer able to dictate prices, the Office profit stream would dry up and so would Microsoft’s ability to finance an attack against Google’s core business.
This, I surmise, is the context for last week’s Google Chrome OS announcement -- and for a rumored Microsoft event this coming week.
With this in mind, let’s look at Google’s pronunciamento.
The content, or lack thereof, timing and manner of the Google Chrome OS announcement raises many questions. Are we dealing with a mere repackaging of a Linux “distro” (the tech argot for a given combination of Linux modules, versions such as Ubuntu or Red Hat) with Google’s own Chrome browser, still “evolving”? Is this the creation of Google markitects, or is it the beginning of the end of Microsoft’s dominance, the lifting of the yoke, at last?
I’d love to think we’re dealing with an epoch-making announcement, with the beginning of a post-Windows/Office era, but I have doubts. Let’s see what those are and why.
First: The OS announcement. Google announces a new operating system focused on netbooks for the new generation of Web-based, Cloud applications. (It will run on PCs as well, we’re told). For processors, it targets two architectures: x86 and ARM. The first designates microprocessors made by Intel and its token competitors. The second powers today’s smartphones from RIM to Apple, from Palm to Nokia. (We’ll discuss Intel’s position/reactions some other time.)
The components of the “new” OS are a kernel, supposedly a Linux derivative, and Google’s Chrome browser. The whole thing will be open-sourced, free to hardware makers. Speaking of which, Acer, HP, Asus, Toshiba as well as Freescale, Qualcomm and Texas Instruments have pledged allegiance to the new operating system. On the software side, Adobe joined in. That’s about it for content. I’m sparing you the full homily about making netbooks and computers run better and the like: “We hear a lot from our users and their message is clear — computers need to get better.” Obligatory statements of principle, would you like like an OS that makes your machine slower, but annoyingly devoid of hows, of specifics showing what measurable improvements will accrue to users. In particular, Google hasn’t offered explanations of the difference between taking a nice Linux version, Ubuntu comes to mind, adding Chrome, the browser, and running all sorts of neat Cloud-based applications such as Google Apps, Zimbra or Zoho, to name a few. How Google’s OS is different/better from/than what we can install on today’s netbooks isn’t stated. But, wait, Chrome the browser runs or will run on existing operating systems, Windows, Mac and Linux; what will be different for the user and for the application developer, besides the obligatory promises of snappier performance? I’ll offer some speculation in a moment. For now, we’ll just note Google offers no hard information regarding the difference between Chrome, the browser, running on existing OS and on their own.
Second: Android. Google has developed an operating system for smartphones, one that is getting a strong following with handset manufacturers and applications developers. The price is right, free, there is a mechanism for developers to sell their wares to users, and there have been insistent rumors of Android migrating to netbooks or to what Intel gingerly calls MIDs, Mobile Internet Devices. By the end of this calendar year, there ought to be a dozen or more Android-based handsets in the US, Europe and Asia. Android isn’t yet a threat to Apple, RIM or Nokia, but it has, in effect, crushed Windows Mobile: how do you sell against free and better. The same fate awaits a gaggle of of Linux-based smartphone developments. So, we have Android off to a promising start in the next wave of computing, smartphones, into the really personal computers, and also looking like a good candidate to power netbooks and MIDs. Why the Chrome OS, then? What’s the difference between the two? Google got Andy Rubin, Android’s chief, to make vague-sounding statements, but no specifics. Perhaps we can help by looking at Apple. They have OS X running on Macs and they have another version, ported to the ARM processor family, powering iPhones and iPod Touches. There are commonalities, but the UIs (User Interfaces) are different, the application models, the development kits are related but distinct. Could this be what Google had in mind? I find the lack of clarity on such an obvious issue rather disconcerting -- or telling.
Third: Google Apps. Of course, the Chrome OS will run Google Apps -- as well as any and all compatible third-party applications. As discussed in previous Monday Notes (How Do You Compete with Free, Google Descencds from The Cloud and Microsoft Mesh caught between the desktop and the cloud) , the new Cloud Computing Order means applications running both on-line, driven by software on servers in the Cloud, and off-line, when the Net connection isn’t available. HTML 5, an upcoming version of HTML, will standardize (we can always hope) how this is/will be done. In the meantime, we can look at examples such as Google Gears augmenting Gmail. There, I work off-line, read my mail, reply to it; everything synchs when I reconnect. New mail downloads into the local cache, mail that was sitting in the outbox now takes off.
There is, however, a fly in the ointment: when I click on Contacts, I’m told they’re not available off-line. The on-line/off-line function set isn’t complete. There are other issues. For example, Google’s excellent Reader (used to fly through RSS feeds and blogs) offers a neat feature: press ‘e’ when reading a post, a panel comes up and lets me email the article. It auto-completes addresses like Gmail, it uses my Gmail contacts as I'm logged on my Google account -- but not quite: some addresses are unavailable, the rest are presented in an order inconsistent with the one in Gmail. Strange. Let’s hope it’s temporary. (I’ve heard there is an internal Google project code-named “Contacts Don’t Suck”...)
These kinks point to the challenge: even for a company as moneyed, as staffed with talented people as Google is, making real Cloud Computing work is hard. Without leaving Google Apps and in an attempt to deliver the promised of speculation: Could Google do anything “special” by owning both ends, the client OS and their own incredibly sophisticated Cloud infrastructure, something like 1.5 million servers running their own secret sauce? Probably.
They could do it nicely, publicly, by implementing selected browser functions deeper into the lower layers of the OS kernel. The results would be higher speed and, possibly, additional functionality. Data storage comes to mind. Today, to manage local data storage, browsers use something called SQLite . New, more ambitious mixed-mode applications will manage their more complex data structures, and do it faster, with a stronger data storage engine built into the OS. Or Google could use the tie between their servers and their client in a more “exclusive” fashion. This is done using hidden, private APIs (Application Programming Interfaces), OS functions only available to Google’s “special” applications. This is a time-honored way to toy with one’s competitors on the same platform: everyone (IBM, Microsoft, Apple...) does or has done it at some point, with mixed results, legal or otherwise. But it’s a possibility, Google could create a tightly-coupled set of mixed-mode Cloud Computing apps, with features built on private APIs, OS functions.
Beyond the word “free”, not much is said of business models spawned by the new OS. As it already does for Google Apps, the company would make basic services free and charge for the more robust versions of applications, of services, including storing our data in the Cloud, a must if you really want to be a serious Cloud Computing player.
Fourth: Timing. This where where explanation and speculation connect.
Remember Google Wave? Not many people do this week. Google Wave “is” a nebulous Cloud Collaboration platform, or concept, announced by Google on May 28th, 2009. Is it one of these announcements where the product of the announcement is the announcement -- and nothing else? We don’t know yet, it’s too recent. But what we do know is Google Wave’s announcement was exquisitely timed to take the media spotlight away from Microsoft’s coming-out party for Bing, their third or fourth attempt at Search, at gaining on Google. (Read what David Pogue, the NYT’s tech guru has to say , it’s quite positive. Microsoft seems to be succeeding by renouncing their old ways, their usual slavish “homage” to Google’s look and feel...)
There are rumors Microsoft will announce some kind of Cloud-based Office/Exchange this coming week. Is Google’s embarrassingly hasty, thin, vague announcement an attempt to steal Microsoft’s thunder, again? (Microsoft was/is never shy about using such tactics.)
Fifth: Speculation. What could Microsoft announce?
Going back to Outlook on Windows, we know it’s the grandaddy of mixed-mode apps. Using Outlook in Cache Mode, I have everything at my fingertips, Calendar, Mail, Contacts, even when disconnected. Whatever I do locally, whatever happened in the server all get reconciled, synched when I reconnect. Now, imagine Microsoft telling us, Look, we’ll get you the benefits, the look-and-feel, the local storage of Outlook/Exchange, but without the local Outlook client. Just use our browser (perhaps a new, more modern one) and our Cloud services. We’ll do the same with every other Office application. Even with Groove, the more real (than Google Wave) collaboration application Ray Ozzie, our Chief Software Architect, brought with him when he joined us.
You pay us nothing for the Basic version, and anywhere between $10 and $20 per “seat”, per month for the real thing, for the Professional version. More if you store a lot of data, always synched with your PC, laptop, netbook. If they’re cheeky enough, they could welcome Mac users in their embrace by issuing a Mac version of their new browser. After all, Mac pushes Windows versions of Safari or iTunes.
Again, this is speculation. We know: over the years, Microsoft has announced a number of products, initiatives, services that went nowhere. But if Microsoft does indeed succeed in building a set of Cloud Office services, or if it merely makes a credible announcement, we’d have an explanation for the timing and content of Google’s moves this past week. Microsoft’s Cloud Office could be a paper tiger or, if real, be toothless. But if it does end up working, as defined by a normal customer, this could be the rise of Microsoft 2.0.
Looking in the other direction, Google might actually be on its way to providing a really good combination of a small, fast client-side OS with a solid set of everyday applications in the Cloud, tightly coupled, working well in mixed on-line/off-line modes, with our data on safely stored on their servers and synched to the client (netbook, MID, PC) cache. In this picture, Google becomes Microsoft 2.0.
Three more thoughts before we go: time, money, asymmetry.
Time: This much-heralded move from desktops to the Cloud will take time, lots of it. Think Big Enterprise IT: How would you like to risk your CIO (Chief Information Officer) career on getting rid of your Exchange servers and Office software on PCs -- and putting all your company’s email, calendars, contacts..., in the Cloud? Individuals and small organizations might welcome such a move, for Enterprise customers it entails large risks, the move will be very, very cautious, slow. Five years or more.
Money: 100 million users, $10/month/user = $12bn/year. This is the order of magnitude we have to keep in mind when thinking of Office in the Cloud, for Microsoft, or for Google. It could be twice as much, perhaps, think how $240/year/user feels, but not more.
These are companies with 2008 revenues of $60bn, for Microsoft, and $22bn for Google, the latter growing a bit faster than the more established Redmond company. In the timeframe just discussed, will the Office in the Cloud revenue stream be a real game changer for either company?
Asymmetry: For each contender, dominating the Office in the Cloud market yields very different outcomes.
If you’re Microsoft, you win playing defense, you protect your Office franchise, at a cost: possibly lower Office/Exchange revenue. This is a better outcome than clinging to today’s desktop hoping the fashion will pass and losing everything to an initially “inferior” interloper.
For Google, you win by playing offense, you gain additional revenue and you keep users spending time on Google properties, exposed to Google-managed advertising.
Time to quit, we’re getting too strategic. —JLG
Here’s what I think its taking place: