All of the stats around the split between XP and Win7 conceal a truth. Large enterprises are struggling massively to muster the energy to make the transition. As I talk to colleagues across multiple industries it is amazing how many big firms are saying:
1. They won’t complete the migration until 2014 or 2015
2. The y fully expect that they will be paying Microsoft for extended support for Windows XP and Office 2003
Microsoft staff will admit this, and that the adoption of Win7 is skewed to consumers or smaller firms. There are even rumours that Redmond may extend the EOL (end of life) dates for XP and Office 2003.
This has nothing to do with the technical quality of Windows 7 and Office 10. Enterprises have simply lost touch with the implications and responsibilities of operating a distributed computing environment. In addition, the accumulated weight of bad discipline around application management, entitlement management for staff, and the eternal battle for the soul of the PC (personal tool or enterprise asset) has made the necessary act of migration to a new OS and productivity suite seem massively onerous.
The inhibitors to speedy migration are remarkably similar at every large enterprise I talk to.
3. ”We don’t have the funding to test and/or remediate applications”
4. “we have too many other priorities and there is no business value to an upgrade”
5. “Isn’t this the accountability of the infrastructure/operations team?”
(sigh) so the irony is that most applications will migrate easily from Windows XP to WIndows 7, and even the bulk of Office macros and plugins will move successfully. But enterprises are the victims of their own processes. Even if Infrastructure folk could quickly move 95% of the desktops to a new image and OS without developer/business intervention, Governance and Quality processes demand that $$$ millions be spent in regression testing. Because LOB’s have no application management discipline, there are way too many applications to test. Even worse, they have allowed staff to deploy thousands of small scale productivity applications which they will not abandon, and thus must also be tested, sometimes remediated, and packaged for deployment.
It is a wholesale abandonment of accountability. If information technology is a core business asset, you have to care about the ongoing viability of the platforms you operate on. Ever since the Vista debacle, the upcoming end of XP has been visible in the news. There is no excuse.
The short term reality is that a lot of firms will end up spending a lot of money twice.
6. They will pay Microsoft $$ for extended support for XP and Office ($1Million the first year, $2M the second year, and so forth)
7. And they will still have to spend the money for testing and remediating applications
Such is the reality of living dependent on the Microsoft product lifecycle.
Even more ironic, they will yell at we poor Infrastructure folk about the fact their staff do not have the latest version of MS Excel at the same time they won’t spend the effort to test and migrate the required macros and plugins that would allow them to use the new version. Dante might allocate a special circle of hell for those capable of this logic (smiley)
I have no doubt firms will stagger through this. But I’m hopeful lessons learned will result in some future changes to the IT disciplines underpinning distributed computing.
8. We need to change the application paradigm. There need to be far fewer enterprise applications and they need to be accessed by far fewer simpler, standards based clients. A browser and HTML5 needs to be the answer for almost everything.
9. We need to radically reduce the number of applications IT has to manage on a desktop. Browsers help, but BYOD (bring your own device) is the answer. Employees need to bring/manage their own device, IT will partition that device via “virtualization techniques” and
a. Manage a radically simpler enterprise application footprint on the “business side” of the device
b. employees can install all those productivity applications they claim are so important (really, Quicken is a business tool?) on the employee side of the device and devote their own energy to maintaining currency. Enterprise IT will be freed from teh nightmare of massive, undisciplined complexity.
There you have it. Mike’s predictions. The nightmare of the WIndows 7 migration will promote three key directions I believe in
- application simplification via HTML5
- Desktop virtualization
- BYOD
Amen. There is truth in every sentence here. Another major influencer in your predictions is also the 3 year product cycle from Microsoft, even adopting “n, n-1″ version control means that governance would dictate a 6 year repetition of this type of project….something major enterprises won’t be able to justify, or achieve.