Does the phrase “data migration” send chills up your spine? Would the fear of moving content from your old website to a new one hold you back from pulling the trigger on a website project?
If you nodded “yes” to either of those questions, you’re not alone.
The topic of data migration can be scary for marketers. Simply put, trained communicators are not database jockeys, and the prospect of ones and zeros flying back and forth can cause discomfort. But comfort level aside, should the fear of data migration warrant the keeping of a crummy website?
What can go wrong?
Everything, right?! There’s a chance that data could get lost, or content could end up in the wrong places, causing countless hours of aggravation for the marketing department…
Professionals who do this all the time are well practiced in the art of handling the various challenges associated with data migration. But how can marketers be assured that their migration won’t go awry.
How to do data migration
Data migration is not rocket science — or magic. At its simplest, it can be summed up as matching database fields from the old website, with the new. The actual act of migrating data (also called “data mapping”) can vary in its level of difficulty, depending on the condition, structure and size of the firm’s current website database. But irrespective of how old your website is or how it was built, the basic steps involved are the same.
Here’s a simplified version of the process that we use at Great Jakes:
- Analysis: The first step involves requesting a “data dump” of all the text content of the website and of the headers for each data table. We analyze the data to determine how much of the migration can be automated.We also investigate whether it would be more practical and/or cost efficient to not automate the migration and instead configure a “data-entry” website to have the data manually moved from the old website into the new. It’s not as hard as it sounds, and it’s not unusual that we end up recommending a combination of automated data-migration and “data-entry” website methods. It all depends on how the old website’s data are formatted.
- Transfer setup: The next step involves planning the “field-mapping” — writing the appropriate scripts necessary to move the data into the proper fields of the new website.
- Migration: A month prior to delivering a finished website, we migrate the data from the old website to the new, using data from a second data dump that contains all of the most current content (text, photos, PDFs, videos, presentations, audio files, etc.).
- Testing: Finally, we rigorously review the data migration results to ensure that everything moved as planned.
Better the devil you know – right?
While the steps outlined are straightforward, data migration is a time-consuming but doable process. Consequently, larger websites with more data will require more time to analyze, set up, migrate and then test.
So, the best way to ensure that everything goes smoothly is to take a peek under the hood. Have a pro examine your existing website. They might find some issues, like embedded tables or miscellaneous image files tucked in strange places. You’ll probably need to make some decisions about how to migrate these items. But most likely, they’ll probably find that the hurdles to moving the content are a lot lower than you might think.
Don’t let those two little words “data migration” keep you from advancing your firm’s business goals! There’s too much to be gained from having a properly conceived website.
Good luck with your migration!
Hat-tip to my friend Nancy Slome, a Web marketing consultant who inspired me to write this post. Nancy provides consultation services to law firms of all sizes on a variety of topics, including data migration.
Content/data migration is usually the most underestimated portion of a website project for the exact reasons you pointed out in your post – it’s overwhelming to take on and can be scary if you don’t know what you’re doing. Seeking expert advice and creating a migration plan are key to a successful project – thanks for the post!
I am sorry, but you did not mention any special technique to migrate data. Analysis, then migration and finally testing are those general steps which everyone knows.