Whether you’re migrating a single application or company-wide infrastructure, there are certain protocols to follow when performing a migration. Not every type of migration has the same set of instructions, but there are common reasons why migration fail. Here are a few of them — and how you can avoid repeating history.
1. You didn’t thoroughly understand your legacy systems
In a perfect world, you’d be migrating from an old system that was impeccably documented, from initial setup to upgrades to emergency patches and beyond. Perhaps in this fantasyland, your entire team has been involved since the beginning, so you can rely on your own experience, plus this documentation, to determine the perfect migration path.
Back to reality: You’re most likely dealing with legacy systems that weren’t well documented or even poorly implemented. Initial decision-makers may have retired or moved on. It’s up to you to gain a thorough understanding of your legacy systems in order to develop the best migration plan.
Don’t neglect the details. If you’re performing a data migration, do you understand the quality of your existing data and what challenges poor data quality might produce? If you’re migrating from on-premises systems to Office 365, have you done an audit to ensure there are no existing incompatible versions of Office or other applications actively in use?
2. Your requirements can’t be met by the new platform
You may be able to articulate the intrinsic value of your proposed migration, but if you aren’t confident that the new operating system or platform meets your organization’s specific needs, that intrinsic value doesn’t mean much. Engage other lines of business or other subject matter experts and involve them in the entire migration process.
Migrations are rarely an “all or nothing” solution to a problem. Often, you are trading one problem for another, and it’s up to you to understand what’s required versus what’s “nice to have.” For example, if you’re moving your email and file storage to the cloud, you may be reducing mailbox size and storage capacity for a subset of employees who are reliant on the current capacity.
The pre-migration discovery process presents an opportunity to understand this problem and find a solution (e.g. is it a business requirement that these employees have more storage, or is it something that can be solved with a different process, like teaching new user habits or automating file cleanup?) before kicking off.
3. You didn’t plan and test enough pre-migration.
If you fail to plan, you plan to fail. Like many old adages, this one is true and crucially important when it comes to a migration. If you’ve taken the time to understand your legacy systems and ensure business requirements are being met, you are thinking like a planner.
Continue with this mindset and create a comprehensive migration plan with a realistic timeline that includes ample time to test. Your plan should incorporate the analysis of existing systems and comparison of business requirements and capabilities that you should already be doing.
Make time for validating and testing throughout the process, so you can discover any potential issues before your migration is complete. Testing requires extra budgeted time, and often the involvement of extra resources from the company, so make sure it’s in the plan from the start.
4. You didn’t communicate well enough on a wide-scale level.
Failed migrations are often due to reasons that could’ve been discovered ahead of time with effective communication. It’s key to communicate well at all levels during the migration process, from executives, to IT teams, to end users. If you feel that you and your team don’t effectively understand the business requirements coming down from leadership, it’s time to communicate!
Conversely, it’s important for you to communicate to end users what to expect before, during, and after the migration. Will there be any downtime? Will any day-to-day systems look and act differently? Let the larger team know what to expect and you may find you learn important information in return.
5. You didn’t train your users.
While it may not be considered a migration failure per se, you don’t want your newly-migrated systems to face adoption challenges. Beyond simply communicating what can be expected, you may need to train your user base on the systems implemented during the migration.
Depending on what systems were migrated, the changes may be unnoticeable — in this case, simple notifications are fine. But if there’s any process change for users, from the way they log in to their machines and beyond, you should provide documentation and training. If you run an IT team or department, members of your own group may need training, as well — both broadly and in new specializations that may arise post-migration, such as this AWS Cloud Practitioner course.
Ensuring a smooth migration from start to finish is not a simple task, but it is a worthwhile goal. As with any challenging goal, reaching it is more likely when you begin with smart and deliberate planning. Your migration will likely impact the business substantially, so it’s important to understand the requirements, involve the right people, and communicate the change. Migrations can fail for many reasons, but if you avoid these pitfalls, you’re more likely to succeed.