The Timely Death of Packaged Applications

[#1]

The Chief Information Officer of a company, as you well know, typically makes long-term expenditure decisions regarding information technology; and for companies without a CIO, the CFO has the final say. Daily, or monthly, operating expenses are usually left to another department. Essentially, that’s what the following piece is about: Your job as a CIO/CFO may get a bit less burdensome within the next three years, as a result of an evolutionary shift in the architecture of applications that now appears all but inevitable.

Over the next three years, all of the applications you use for everyday business productivity — including and especially Microsoft Office — may be delivered to you through the Internet. There’s no disruption going on here, just common sense: Continuing to maintain PCs — separate boxes with huge hard drives, each storing unmanageable caches of redundant data — is becoming impractical.

Consider for a moment: Why do we continue to use sophisticated client-side operating systems?

• They manage and maintain the seats for software licenses. With business users commonly accessing their networks with more than one device daily, it’s making less sense for businesses to maintain multiple seats per user, or for vendors to enable them in their licenses.
• Operating systems manage the local hardware attached to PCs. The most obvious example is the hard drive, whose size has ballooned mainly because of the rapid explosion in media files and the gargantuan size of local applications, most notably Microsoft Office. If applications were delivered and managed from a centralized image and only used remotely, and if media files could be streamed from centralized sources without being replicated, then you wouldn’t need colossal hard drives to provide exactly the same functionality. And since plug-in devices are becoming standardized around USB, it’s more feasible for those devices to be managed by client-side firmware rather than by the OS.
• Sophisticated operating systems provide a central point of security. But that’s the problem: Bigger and more mobile PCs become bigger and easier targets for more clever malware. And since malware is becoming less normalized, more multifarious, and more social, security is more a matter of educating users than applying fixes, the cost of which is the cost of education itself.

One of Microsoft Windows Server 2008’s greatest innovations was the ability for administrators to deploy images of installed operating systems and fully customized applications to users, rather than running the installation scripts locally (and mitigating the inevitable damage). But with the bandwidth cost declining so rapidly, administrators and system architects are rightly pondering the following: If the only reason for deploying images of functionality is so users can have applications locally, when localization is becoming the problem, then why not find a way to deploy the image centrally and distribute it virtually instead?

It’s making more sense for businesses (and sooner than you think, for consumers) to maintain their users’ applications, settings, and identities (what we still call desktops) remotely. Today, enterprises install desktops remotely, and then admins use Remote Desktop Access to help users maintain them. Why not install desktops locally instead — or perhaps more to the point, why does a desktop have to be a thing? The ideal of the virtual desktop is effectively to relocate what we consider to be the user’s computer to a centralized location where it can be more easily managed, and enable it to be streamed to whatever system or device the user happens to be logged onto at the time.

When Behavior Changes the Environment

Soon, though, the question why not very quickly becomes why. Once applications are relocated to “the cloud,” the nature of the work these applications perform changes.

Take PowerPoint for example, and the work environment around which PowerPoint revolves. Many first generation consumer-grade cloud services are based around storage relocation. Several borrow their value proposition from the world of iTunes and MP3 players, such as Amazon Cloud Drive and DropBox for cloud storage, and YouSendIt for cloud-assisted transfer of e-mail attachments. In the same way Apple estimates how many songs fit on an iPod, some services use an estimated average PowerPoint presentation file size to represent storage capacities. With 10 GB at your disposal, say some, imagine how many PowerPoint decks you can share!

Meanwhile, the second generation of cloud services is already under way, with providers such as Cisco’s WebEx offering live whiteboards and HP’s Halo videoconferencing. When a workforce is brought together directly with live communications, and collaboration takes place in real-time rather than by e-mail, imagine how many PowerPoints you no longer have to save, share, revise, or re-version.

With more broadly available bandwidth enabling functionality to be more efficiently relocated from local storage into the cloud, what is likely to change is users’ behavior. That will trigger a change in the type of functions that applications provide, especially with respect to collaboration and workflow. To date, the way people “collaborate” around documents is one step removed from the days of copying files to floppy diskettes and circulating them in inter-office mail envelopes. A surprising number of companies in the business of producing information continue to “collaborate” by attaching multiple, separate versions of documents to e-mail messages. For them, the versioning system of Microsoft Office applications (if they’ve discovered it yet) is a convenience. If a new business adopts a bold, new application on the order of WebEx conferencing, then yesterday’s convenience becomes today’s doorstop.

The keenest observations these days about how user behaviors are affecting the nature of applications as they move to the cloud are coming from the security sector. There, analysts evaluate new trends in user behavior in order to predict how malware authors will exploit those trends next.

Ian Moyse, who directs Europe, the Middle East, and Asia for cloud security provider Webroot, conducts his own experiments with everyday users at security conferences worldwide. “I’ve asked a lot of end-user audiences a simple question: ‘How many of you get your e-mail on more than one device?’ Hands go up,” Moyse tells us. “‘How many on more than two? Three?’ I’m commonly now getting to four and five, and I’m still getting hands in the air. So you connect this to your e-mail, anywhere, on any device — your iPad, your PC, your BlackBerry, your iPhone. You don’t care about how it works, it’s not complex,” he says.

Ten years ago, it would have been difficult to access your e-mail from so many devices.  “Now, it’s so easy,” Moyse continues. “Home users expect to get their e-mail on the mobile phone from their Gmail or Hotmail account. When they come to work and they can’t do it, it doesn’t make sense. So users are coming into business who have been educated to assume things are available on the Web. They take it for granted that the power is there, and that it works. And now there’s an expectation on business to consumerize the delivery of service to the users, to be competitive and deliver that quality of service from IT.”

The consumerization trend to which Moyse refers can be found in online apps such as Google Docs. There, multiple users can see the changes each other makes to a document, as those changes are being made. It’s the type of feature more users are asking to see in Microsoft Office, and which Microsoft may not have been all that pleased to see Google providing first, with an Office add-on called Google Cloud Connect.

With users’ expectations rising and Microsoft’s competitors — for the first time in over a decade — applying serious pressure, Moyse believes that the changes to the entire software ecosystem precipitated by cloud services will be gradual, not sudden, and in four interleaving stages:

1. First, new vendors will enter the applications space, knowing that for the first time, deployment cycles are no longer an issue. “Deployment’s gone,” Moyse says, “the focus is on development.” This changes the pace of innovation from Microsoft’s three- to five-year phase to whenever someone like Google is prepared to issue a feature like Cloud Connect.
2. Software distributors — a term that already sounds prehistoric — will continue their transformation to aggregators. Specifically, as these distributors become the servers of functionality, they will also be responsible for provisioning: making sure that services are available to customers, and are provided according to their requests. “Their model is changing,” remarks Moyse. “It isn’t now purely about moving a license from one location to another, and billing.”
3. Software resellers (historically distinct from distributors, although the distinctions may disappear) will transition to a point-of-contact role. Resellers have typically provided installation packages, and perhaps for some customers, installation services. “It isn’t going to be about installation anymore; it’s going to be about configuration and making sure these three components work together,” says Webroot’s Moyse. “It’s [about] becoming a true value-add to that customer’s line of business.”
4. All this then triggers what some perceive as a long-awaited alteration, and others as a sudden transformation: Businesses and enterprises will shift at least some of their reliances from their own IT departments to these services.

So what are the consequences? Without installation and maintenance, does your IT department have anything left to do? Even security can be effectively delegated to experts in the cloud (which Moyse, the Webroot executive, happily points out at every opportunity).

This is going to change how the financial model works, according to Moyse. “First, the services sit on the operating budget line rather than capital expenditure,” he tells us. “So the decision-making process for procuring software changes to something more like leasing company vehicles or apportioning travel expenses.” For bigger businesses, this could mean the decision maker changes, from someone who reports to the CFO to someone who reports to the COO.

Let’s repeat that: Points of contact for decision-making are relocating. “A lot of customers’ cloud decisions are more heavily influenced by the business decision makers — finance, and the service line manager or marketing manager — than was done previously, when it was typically a technology decision,” says Moyse. “Previously, for a marketing director to put an application into the business, it was very difficult to do without involving IT, and without IT doing it for them. Right now, with a credit card in their hand, [the marketing director] can conceivably turn on a cloud service, give access to the marketing team, and have it up and running by tomorrow morning – and be using it without IT even being aware of it.”

When the Environment Changes Behavior

If both the decision making process and the decision maker changes, then Microsoft has a daunting task ahead. Historically, its value proposition to business customers has been based on platform leveraging. For instance, you installed Windows because it was the best platform to run Office. You ran Office because it made the best use of Exchange. You used SharePoint because it simplified the workflow for Office. Now, Microsoft has to justify Office as a brand in front of an altogether different customer.

Which makes its Office 365 program, announced last October and put into beta test in June, all the more significant in determining whether Word and Excel will share the same fate as Multiplan and VisiCalc. The fact that Microsoft is moving with some trepidation toward this new licensing model is lost on absolutely no one. Office 365 is not a cloud service, even though cloud services, to a limited extent, will comprise Office 365. But it does represent Microsoft’s best effort to date to adopt (perhaps “embrace” isn’t the right word just yet) a per-user subscription model, with a complex scheme for enabling current Business Productivity Online Suite (BPOS) subscribers to “bridge,” to use Microsoft’s term, to Office 365.

The Office 365 suite encompasses three of Microsoft’s other main platform brands as fully hosted services: Exchange for e-mail, SharePoint for collaboration, and the newly renamed (and better) Lync, formerly Unified Communications, for supplying digital points of presence (though not with PBX functionality and the ability to hook up with the phone network, as with on-site Lync). The value proposition is in transitioning customers from the traditional Client Access Licenses (CALs) to per-user service agreements that encompass any or all three of these hosted service brands, plus Office.

But it isn’t really “hosted Office” as some folks expected. Rather, it’s a per-user subscription plan that may include the conventional packaged Office Professional Plus applications, plus access to Office Web Apps — which are not fully-functional equivalents of Office apps, and which have actually been made free to consumers anyway. While it’s confusing, the new plan may end up being less expensive for some businesses over time.

Here’s how the bridging concept works: Many Microsoft customers may have CALs that have yet to expire; for example, the company’s Enterprise CAL Suite already incorporates Exchange, SharePoint, and Lync along with Forefront Security and Windows Server, which are not part of Office 365. So for these portions of the existing CAL that don’t overlap, Microsoft offers customers a swap for a “bridge CAL” agreement that picks up the remaining time from the old CAL, and is coterminous with the new O365 service agreement. Customers have their choice of tiers for O365, including a lower “E1” tier that excludes per-user licensing for Office, a middle “E2” tier that includes Office Web Apps, and an “E3” tier that includes limited per-user seats for Office. (Supposedly when one’s subscription expires, she is not authorized to use the product, but it remains to be seen whether Microsoft will invoke “software de-activation” for this purpose. An “E4” tier with full telephone functionality for Lync is expected later in the year.)

This gives decision makers (now including a marketing manager or two) something more like the mix-and-match system of choices they’re accustomed to when selecting services. Webroot’s Ian Moyse sees this as one development which could trigger the long-awaited exodus by Microsoft to a full service model, and a validation of the trends being measured everywhere else in the market. But he feels it won’t be a sudden shift, which might suit Microsoft just fine. Instead, it could be an exodus so gradual and so effective that once it’s done, folks might not miss the old world very much at all — as when he holds up a 5 1/4" floppy diskette at a conference and gets confused stares from onlookers wondering what it is.

“It wasn’t so long ago that we took [the floppy diskette] for granted; and now it’s phased out and we don’t even notice things have replaced it,” Moyse says.  “It didn’t go away in one day, but slowly, the same way tape went to CD. Things evolve and you don’t notice them.”