5 key principles for a successful content migration

5 key principles for a successful content migration

Digital transformation journeys can be exciting for enterprise organizations looking to evolve their digital ecosystems. But they can also feel daunting. Though digital transformation promises to bring new capabilities for delivering more effective and engaging experiences to customers, the process can also disrupt every part of a business with the added pressure that everything should operate normally. The prospect of moving hundreds of thousands of digital objects critical to your business can create anxiety for many marketing leaders about how to get started.

For organizations that manage large amounts of content, transformation is especially intimidating because that content needs to maintain its integrity while it moves out of legacy systems into new platforms.

Digital transformation projects that include implementing a more advanced content management system (CMS) like Adobe Experience Manager will include a migration phase. That phase will be one of the more complex phases of the project. Migration is typically executed during the final stage, which adds even more complexity to the project’s overall success. To make the journey worth the effort, it’s essential to plan for all phases of the migration to ensure the transformation ends with users adopting all the new capabilities.

Every digital transformation should include a content strategy that considers the current and desired content lifecycle (and that strategy also needs to align with immediate and future business goals). No matter how challenging your migration may seem, following these five guiding principles can help ensure success even with the most complex migration efforts.

1. Inventory: Locate what needs to be migrated

Performing a proper inventory analysis of the content that needs to be migrated can be an illuminating exercise for any enterprise. Surprisingly, many organizations do not have a clear or confident understanding of how much content they own. The first step in setting up an inventory is understanding the level of effort needed to create the inventory analysis. For some organizations, this may be a simple task — but for others, it may require a lot more effort, including discovery, planning, and preparation.

Understanding how and where content is stored will help you generate your inventory lists. Creating these lists may be as simple as running a web crawling tool on the site. The source systems storing the content may also have reporting tools to extract the list of objects living in the system as well. For other platforms, it might require custom scripts to query systems in order to generate the required data. When doing this, it’s important to extract as much information as possible that’s associated with the content. This includes the name, location, all metadata, and any linkages that reference the content.

Establishing your content inventory is a critical first step to understanding the scope of all content that needs to be transformed and migrated. Moving forward without this inventory means that all subsequent planning might contain unrecoverable gaps and blind spots. The inventories should be organized in a manner that allows for a deeper analysis, like a spreadsheet or data visualization tool. This will enable the ability to optimize the items by reordering, editing, adding, and removing information where necessary.

Taxonomy and metadata — optimize content organization

2. Taxonomy and metadata: Optimize content organization

While not required for all migrations, a digital transformation project is also an opportunity to evaluate and optimize how your content is organized. Consider carefully what your goals and objectives are for revising your taxonomy. For example, the aim may be to simplify and reduce complexity or perhaps expand and make room for more content. Regardless, there should be measurable methods in place to determine the effectiveness of the changes made to your taxonomy.

If you decide to revise your taxonomy, remember to make it realistic, simple, and clearly understood. All the content listed in those inventories will need to be mapped to the new taxonomy. A lot of frustration can be avoided after creating a sophisticated taxonomy only to realize that it will take months to map the current content to the new taxonomy.

This is also a great opportunity to improve content metadata to achieve findability and SEO goals as part of the digital transformation. For example, there may be several high-value web pages that lack proper descriptions. Adding descriptions will improve their SEO scores and make those pages easier to find. This is why it’s so important to ensure that all metadata associated with the content is included with your inventories, as that will allow you to understand and quickly identify where there are gaps that need to be filled.

Filling in the gaps — or more specifically, adding missing metadata to your content — requires careful prioritization. And just like revising taxonomies, the exercise of adding the metadata needs to be practical. A cost-benefit analysis should be used to determine if the effort will increase the value of the content. Business metrics like content velocity, content reuse, findability, searchability, and governance optimization are all positively influenced by incorporating more robust metadata.

There are three opportunities to add rich metadata to content during a transformation:

Each choice comes with its own set of advantages and disadvantages. Ideally, the content updates happen in the native system, allowing the organization to use familiar tools and processes to make the updates. When this isn’t possible, the other two options will need to be assessed to determine the path of least resistance. Just be sure that there are protocols in place to ensure it gets done. We often see our customers treat this last step as an afterthought and assume mistakenly that it will happen automatically over time.

3. Clean up: Start with a clean slate and don’t migrate what you don’t need

Cleaning up content is comprised of two key steps that require careful attention. First, you need to prune your content and remove inventory that is no longer necessary. This may include old content that is no longer accurate or relevant for your intended audiences. Second, you need to cleanse your data of any malformed, illegal, or illogical terms and characters. This is an opportunity to introduce better naming conventions and standards for terminology. It’s important to ensure that file names and associated metadata are also sanitized.

content that shouldn't be migrted, old content, low value content, malformed content, illegal content, illogical terms and characters.

There are several ways to determine what should and should not be included in the migration, but the two main factors are age and relevance. Content age alone may not be enough as it should also be coupled with its value and purpose. It’s possible to have evergreen content that is as old as the first platform release that still holds relevance to your customers.

4. Migrate: Out with the old and into the new

The process of moving content out of one system into a different system is unavoidably complex. One crucial component of a successful migration is to break it down into parts that are repeatable and easy to validate. A single piece of content may contain any combination of data types, including text, media, files, metadata, and code. The content may appear as one thing to the customer like a blog page with paragraphs, images, and a video with an accompanying transcript and a link to a downloadable file. Each individual data type may live in a different location with its own set of rules on how it can be accessed. The text may live in a component in the content management system, the image may live in a DAM that is delivered by a CDN, the video may be on an entirely different platform like YouTube, and the downloadable files may live in another repository. The code that controls the presentation may leverage another programming language outside of the CMS, like jQuery or React. All of these variables need to be accounted for just in the extraction alone.

Before executing the transfer, a sample dataset should be used to validate the process from end to end. The sample dataset should include information from all sources to ensure every variable has been accounted for before the transfer happens. The sample set should also be large enough to generate a strong estimation of the required duration for all the content to migrate. This includes determining how long it takes to get the content out of the current system and transferred and fully processed into the new system, as well as understanding how long it will take to reconcile the delta of what changed while the transfer took place.

Those transfer estimations can be used to coordinate the logistics of performing the complete migration. This is critical to ensuring the business is still operational during that time. No matter how long the migration takes — hours, days, or weeks — changes made during the migration need to be addressed. At some point, content will need to be frozen to maintain the integrity of the transfer. If content is added, edited, or removed in the legacy system during the migration, those changes will need to be reflected in the new system.

A validation method should also be defined to verify the integrity of the content once it has transferred. While it may be impossible to physically check 200,000 web pages and 600,000 assets, there are still steps that can be employed to reach a validation criterion that is acceptable. We recommend two approaches:

  1. Define a validation sample dataset to check thoroughly
  2. Define another larger dataset that can be scanned for partial data

Like all other principles highlighted in this article, be sure your validation methods are achievable.

5. Governance: Add controls to maintain operational excellence

When your migration is complete and your content is flowing though the new system, it’s imperative that an ongoing governance model is in place to keep the new system in an optimized state.

Leverage the new system’s capabilities and enforce governance with access permissions, workflows, and reporting. Create policies and procedures for users to adhere to. Lastly, make sure there is adequate training for all users so they understand how to properly leverage new capabilities using best practices and standards.

Next steps

If your organization is thinking about a digital transformation that requires content migration, remember to be realistic and stick to the five principles: inventory, taxonomy, clean up, migration, and governance. By creating a plan across these five areas, you’ll be able to deliver engaging customer experiences in no time.

Want to learn more about the migration process? Adobe Professional Services can help.

Shawnn Guthrie is a principal business consultant with Adobe Professional Services helping customers leverage Adobe’s content and commerce solutions. Guthrie has more than 20 years of experience leading digital transformation initiatives. He has provided strategic leadership for some of the largest brands in the world and brings a relentless pursuit of excellence to every engagement, ensuring customers realize tangible value with Adobe's solutions.