Last year there was a great deal of focus on Windows XP, as businesses sought ways to move away from the ageing operating system before the April end-of-support deadline. Stories of long, overrunning migration projects were commonplace. What's more, businesses faced significant cost penalties if they failed to migrate before the deadline. The NHS alone was forced to spend an eye-watering £5.5 million (around $8.5 million, AU$10.9 million) to extend its support for another year.
Yet despite the clear need for rapid migration, two months after the April 2014 deadline, over one-third of UK and US organisations either hadn't started migrating or were still in the process of doing so (according to research our company commissioned last year). Many businesses simply couldn't make the deadline.
We're still facing the same problem
Now the next big deadline, the end of Custom Support in April 2015, is looming close on the horizon. This poses a problem for many organisations as our research shows that an operating system migration takes, on average, five months (with many taking longer). In other words, you likely need to have started before now…
The simple truth is that businesses must find a way to radically cut down their migration time. There is also a need to find a way to deal with the accelerating pace of OS migrations in general. Windows 10 was announced at a time when most businesses had not even migrated to Windows 7 or 8 and this trend is likely to continue, making migration a 'business as usual' consideration, rather than a one-off project.
What's the problem with current migration methods?
Put simply, most enterprises simply aren't set up to handle migrations as effectively as they could be.
Many issues occur right at the beginning of the process. Too many businesses fail to take stock of the applications they have, which ones they need, and which ones can be removed. They therefore spend an undue amount of time migrating redundant applications as part of the process, leading to unnecessary timescale and budget overruns.
Another problem is an excessive reliance on manual intervention during the migration process itself. Many businesses still adhere to lengthy manual OS migration approaches, sending workers to computers with install disks in hand. Others rely on physical mail-outs of install disks (which can take months to coordinate), or even shipping machines elsewhere to be upgraded.
Even in cases where the intention has been to handle migration and distribution over-the-network, businesses have run into 'gotchas', often around build processes, that have eaten up time and budget through an unacceptable number of manual interventions. What's more, these processes are often disruptive to 'business as usual', committing the cardinal sin of disrupting the user.
The benefits of zero-touch
Zero-touch is a deployment approach that has been proven to accelerate, automate and reduce risks for large-scale Windows migrations. It places emphasis on high-level automation and optimisation of software distribution.
Typically, in a zero-touch migration, full automation of migration on 90% of machines is aimed for. This can bring cost and timescales down by an order of magnitude. The process also focuses on causing minimum disruption to the user and organisation.
Migrations with some of the UK's largest companies have shown this approach is more than four times faster than competing methods. In fact, it has proven to be the fastest migration process in the industry, having overseen the migration of as many as 30,000 PCs in a single month.
Another benefit of zero-touch comes in terms of staff utilisation. Even in the 10% of cases where PCs require a desk-side visit, it's unlikely that a technician will have to walk through the entire installation process. Most of the time only partial manual processes need to be implemented, freeing up the IT team for more useful tasks.
As a process, zero-touch can also help users. Rather than being forced to adhere to the IT department's schedule, users can schedule upgrades themselves at a time which suits them (with upgrades often scheduled to apply themselves automatically outside of working hours). If users feel the IT team considers the potential impact on their working day, and that their needs are acknowledged, they'll be more sold on the migration process.
But the ultimate argument comes down to cost. Gartner has previously stated that the cost of migrating a single XP machine will generally be between $1,205 (around £780, AU$1,550) and $2,069 (around £1,340, AU$2,660). Zero-touch migration brings this down to only $8.50 (around £5.50, AU$10.90).
What should you do?
Before embarking on a zero-touch migration, the business will obviously need to establish objectives, budgets and timeframes, and run basic hygiene tests on infrastructure. You'll also benefit from engaging with a specialist. Generally, the amount of time and money a clean migration will save will more than pay for itself.
In practical terms, the pieces you'll need in place for a highly automated migration are SCCM, as well as an app survey tool that is fully automated and does not use agents. The business will also need a system for mapping legacy apps to more recent equivalent app versions, and a method of efficient network deployment.
Critically, establishing a process for zero-touch migration should not be seen as a one-off reaction to individual migration projects. Rather, it should be viewed as a way to create best practice going forward, future-proofing against rapidly accelerating OS releases and providing a highly effective route to continually review and optimise the way that you manage applications, software and systems.
- Ambareesh Kulkarni is Vice President of Professional Services at 1E