Want to know the single most important thing to do to optimize your content migration? Think through the steps of handling and evaluating content during the website migration.
Unfortunately most focus exclusively on one single step: How to get content from the old system to the new one. A successful migration requires a further five steps.
Six steps of content handling
Here are the six steps of content handling:
- Sort. Overall what am I going to do with this content (throw it away? move as-is? archive?)
- Place. Where will this content live on the new site?
- Edit. Do the words (or other components) need to be edited by a person?
- Move/Transform. How is the content going to get from the old system to the new one?
- Enhance. Do I need new metadata, and if so how do I accurately apply that?
- QA (Quality Assurance). Who or what will be QA'ing, and what standards will we evaluate against?
If you only focus on #4 which is common, then you're missing crucial steps and making the biggest content migration mistake.
Common problems with content migration
Leave out thinking about the migration until the very end (assuming that it's just about moving content from point A to point B) and the whole project may well grind to a halt when things don't join up and fit together after the "migration." For example, you may find your migration scripts or temporary staff don't know where to put swaths of content, requiring you to scramble to restructure the new site, jettison content, or some other remedy that takes time late in the process (thinking about how content will be placed would prevent that).
A common problem is actually the surprises that arise in relation to quality levels, when for example the editors expected higher quality pages. For example, if the editorial team feels that the resulting pages are not high enough quality, then the project may need to halt to discuss the overall needs of the migration (which could have been avoided if QA had been planned in more detail). This often happens since the context in which the content appears will change during a migration. Instead of simply relying on QA to catch this sort of contextual problem, consider how the other steps will need to handle it. On example is defining rules for where content should appear in the new IA, rather than finding problems late in the process.
Content migration insider tips
Do not carry out all steps for all content items. The point is to analyze your existing content, looking for commonality within your content. For instance, your could opt to treat press releases older than a year differently than new press releases by putting them in an archive section that is plain text whereas the newer press releases have a richer presentation. You could then consider the six steps for these older press releases, where all steps aside from a sampling for QA could be automated (and the sorting step for example would not need to happen for each content item since it's just applying a rule). After looking at all the buckets of your content, you can come up with an overall estimate of the manual effort for your migration.
The process of planning the content handling process is dynamic, since there is a constant interplay between quality and cost. At the beginning you may expect certain quality levels that are too expensive to attain. For instance, I was working with a large news organization's intranet that would require massive manual editing (the words literally had to be rewritten, with nothing technical about it), and we were able to determine that their original vision would require huge manual effort. Instead we bucketed content in different ways to do far less manual transformation. At other times you discover that there is a middle-of-the-road quality level that costs far less but gets 80% of the desired quality.
Learn more about content migration
For more information on the six steps, see Content Handling Process: Asking the Right Content Migration Questions with links to more resources. Or, for a broader look at planning website migrations see Website Migration Handbook.
Very nice article. How about adding a USE section. You identified where you want to STORE the information. What is lacking (IMHO) is the part about who and how users will access that info!
Thanks for the comment. I would say that the use is underlying all of these steps. After all, the only reason to even have content is to be used. So when sorting content for example, the use is an important input (for instance dropping content that isn’t already being used). When deciding where to place content, putting in the right place for maximum effective usage makes sense.
[...] på över 10 000 webbsidor som ska flyttas över till en ny plattform? David Hobbs har skrivit ett bra inlägg om hur innehållsmigrering är en process som består av sex steg: sortera, placera, redigera, [...]