For most today's companies, data migration simply is a must. Systems change, databases change, and so do business processes, operations and whole strategies. It means that people have to refresh their data and transfer it from old storage places to new ones. The ones which better address requirements of today's business. Data migration itself isn't really a complicated process. It's very deeply described in different sources, different approaches are introduced and there always is a methodology to analyze. But even that doesn't guarantee the success.
Why data migration specialists, even if following the best practices, sometimes happen to be wrong? Because they do mistakes which is understandable (but expensive and time-consuming, in many cases also very risky).
Nevertheless, it's not like these mistakes happen because of no reason. In fact, they're likely happen in the same or similar areas and they're a consequence of typical signs of negligence.
Before data migration
Many issues arise before any migration process actually starts. That doesn't sound too good. But the optimistic side is that they can also be solved before they turn into any serious consequences. Thereupon, every data migration has to be preceded with responsible preparation, considering especially the following issues.
Specialists needed
One of the most important and common problems with data migration is the fact it's disregarded by company management, which representatives don't pay enough attention to the importance of data migration projects. They consider it one of many routine tasks which IT employees should be able to pass through without any special help nor support. Unfortunately it's not. Data migration isn't just a matter of IT, it's a process which involves almost every employee across the organization. Therefore it requires engagement not only from IT, but also from experts from other departments, data users, etc.. What else is needed is to allocate powerful system resources to data migration so it could be completed quickly and efficiently.
Unreal scoping and budgeting
Immensity of data migration tasks might be overwhelming at the beginning, but it's necessary to identify and describe all things which will have to be done. Omitting it is a very bad idea, because it doesn't give a proper insight into the scale of action and, also, its cost. It's extremely important not to underestimate the cost of data migration - looking for savings on this stage is suicidal. The same is to scoping - underestimating the number or intensity of things which have to be done makes data migration longer than predicted and, therefore, much more annoying to data users across the organization.
Data inconsistency and lack of management strategies
It definitely isn't easy to keep data quality on its highest level. However, it's required to keep data as usable as possible. Unfortunately, it doesn't happen in many organizations. In most of them, data quality is getting lower and lower in proportion to time which passes since the previous migration or data cleansing. In fact, proper data management strategies and regular data cleansing are the best methods for keeping data quality on the demanded level. The level which, in time of migration, minimizes a number of things that have to be done to prepare data to transfer.
During data migration
Wrong migration strategy
What to do next? It's one of these questions which shouldn't be asked when data migration is started. People who are involved in data transfer have to know very well all steps of data migration and have to know what's next. If the strategy is incomplete or - even worse - it has not been prepared at all, chances that migration will finish with success are getting lower and lower in proportion to the complexity of processes that have to be done anyway.Additionally, there is a tendency to choose big bang migration observed, even though high level of risk it's connected with should be a serious warning to data migration organizers. Many specialists say on the other hand that choosing trickle migrations instead, even though it lasts longer than big bang migrations, is safer option for businesses.
Wrong tools
The choice of tools should be a part of data migration strategy, but the fact is that managers from different organizations don't pay a real attention to tools designated to support data transfer. In many cases, the only tools used for data migration are the ones which the company already owns. They're not designed to real data migration, they're often made to support totally different uses. Purchasing new tools for data migration isn't what many organizations consider right.
No cooperation among teams
Data migration truly is a complicated process which - if has to be done as quickly as possible - requires involvement of people from totally different departments across the organization. If these teams do not cooperate in a required way, the risk of mistakes and - in some cases - counteractions is very high. When two different teams have their own strategies of migration and follow different methodologies, it's almost impossible that everything goes the way it should.
Inflexibility
Good and detailed plan truly is a first step every successful migration starts with, but it not always is possible to keep it alive. The worst one can do during migration is to keep the already fixed plan regardless of changing situation. Sometimes it is a must to introduce some changes and improvements when the work is in progress. Good practice in data migration is to be flexible even though the migration plan is something what shouldn't be completely ignored.
After data migration
Migration lateness
One of the worst things that happen after migration is realizing that it wasn't needed. Or - what also happens, and may be even more common - it's been done too late. The situation in which business operates, is changing constantly and so do business processes, systems, and applications. Waiting too long between fixing migration plan and performing migration makes the presumptions out-of-dated. Thereupon, spending a few days on migration might come out as a waste of time after all.
Lack of proper after-migration validation
The last step of every data migration should be validation of its results. But, what's important, it should be done not only by the ones directly responsible for the process. The real users of transferred data need to confirm that everything is the way it should be. They're the most probable group to identify the issues which may seem not worth mentioning, but turn vital later.
The ones listed above are the most common problems which actually don't have to lead to serious "disasters". It depends on the situation, sometimes data migration is as easy as possible and whatever one does, migration won't fail. However, it's much more often that a single negligence lengthens the migration and increases its cost which aren't neither necessary nor suggested from any point of view.