We all operate within the information age, where data has become one of the world’s most valuable commodities because there is so much potential locked within it. To take more advantage of your data, organisations are looking to increasingly leverage more cloud-based tools and solutions. This often involves migrating on-premises data into cloud solutions or into data lakes. It can be challenging to move this data due to the sheer scale of information sources involved. During a data migration, your business could also expose sensitive data, either when on the cloud 'in transit' to the cloud.
Due to these challenges, there are several myths I have come across about how data migrations work. These myths are caused by misperceptions of the challenges, which mean that expectations of how a migration will happen do not always match the reality. As a result, I have identified a few of the most common ‘data migration myths’ to help you take the risk out of your next data migration.
Myth 1 – Data migration is a one-time task
Myth: Data migration can be done as a one-off task with a clear beginning and end.
Reality: Contrary to popular belief, data migration and clean-up is something that companies must do on an ongoing basis. Organisations, especially ones that have been around for a long time, have numerous old data silos and redundant systems. As a result, it can easily take years to migrate and consolidate them.
Due to the amount of time and effort needed to successfully migrate to the cloud, some companies put it off for years, waiting for technical resources to free up or for a legacy platform to deprecate or worse just dump it in a data lake and paying for it to be stored.
Recommendation: Don’t put off a migration – plan and execute but do a full scope and ensure you allow extra time for unforeseen hold-ups in the process.
Myth 2 – Running apps on the cloud is always preferable to running them on-premises
Myth: Many organisations believe that in moving applications to the cloud, working processes will become more efficient and cost-effective, silos will disappear, and systems will integrate seamlessly.
Reality: Cloud-based systems rarely work without a hitch. In fact, nearly three-quarters (74%) of companies have moved at least one of the apps they moved to the cloud back to on-premise after failing to achieve the anticipated benefits of the cloud-based migration1.
Recommendation: Weigh up the availability, savings and benefits of accessing online apps against their costs and potential security risks. Approach online apps on a case-by-case basis rather than with a one-size-fits-all approach.
Myth 3 – Running apps in the cloud is more efficient
Myth: Cloud services are inherently efficient due to their scalability and versatility.
Reality: Recent research has shown 45% of companies pursuing a strategy to get all their data onto the cloud can over-provision by as much as 55% during the first 18-months and overspend by 70%2.
Recommendation: A key feature of success is knowing what data and services you’ve got before attempting to move it to avoid overprovisioning or moving more data into the cloud than necessary. For this, you can use tools like Microsoft’s Data Discovery or Exonar Reveal.
Approached in the right way, cloud migrations are successful when all the relevant data is identified, prepared, and remediated. Business processes should be optimised before the actual move happens.
Myth 4 – Data migration is a matter of copying data
Myth: Data migration is simply copying and pasting data from system A to system B.
Reality: Cloud shouldn’t be treated as a virtual data centre where the organisation simply ‘lifts and shifts’ the IT infrastructure. Unfortunately, systems usually have different data models which means almost all migrations involve extensive transformations of data simply to get it from one model to another. That is why simply mapping and copying data is not enough.
Recommendation: Consider a migration done correctly as an opportunity to improve data models and data quality as well as radically repurpose old systems. Put in the extra effort upfront to optimise the structure of your data. Don’t be afraid to hit the ‘delete’ button once you’ve identified your duplicated, redundant, obsolete and trivial (DROT) data. This will make your life easier going forward.
Myth 5 – Data migration can be done with just a temporary reassignment of resources
Myth: You don’t need to assign dedicated resources to data migration.
Reality: Data migrations usually demands a permanent investment from organisations for them to be done successfully.
Recommendation: Since data migrations usually involve data integration, consolidation, quality and system upgrades, organisations need to assign dedicated resources to data migrations, especially organizations that have a lot of old data and redundant systems. It should support the personnel and tools as well as fit the budget and executive mandate.
In summary, take a cautious approach to your next migration and be aware of the pitfalls before you start. This normally means adopting a hybrid cloud/on-premises model. With careful planning, you can streamline your data, extract maximum value, and make it future-proof.
For more information, watch the webinar on De-risking your data migration hosted by the Cloud Industry Forum here.
- The Bi-Directional Cloud Highway: Critical Insights into Today’s Cloud Infrastructures (fortinet.com)
- Gartner. (2019), Top 3 Drivers of Cloud Adoption in Midsize Enterprises (gartner.com)
About the author:
Adriaan Bekker is Technical Director of Softwerx. He has worked with Microsoft and at corporate ‘C-level’ clients for over 10 years. This extensive experience, combined with his IT and business degree qualifications, means he is well placed to bridge the gap between technical, commercial and compliance requirements.
Follow Softwerx on LinkedIn and Twitter for the latest updates:
Back to Blog