Forrest Gump’s famous quote “you never know what you’re gonna get” reflects what many of us worry about when it comes to legacy data migrations.
Moving to a new system means analyzing the way data is defined between old and new, and translating your legacy information into the new system effectually. Even migrations that seem black and white on paper can run into obstacles, so planning is crucial. The most successful legacy data migrations are detail-oriented and ask the right questions. What are those questions? What are the details you need to consider? Here are 6 tips for successful legacy data migrations.
1.Choose the Right Method
There are two ways to approach a legacy migration: migrate everything into the new system as-is and wade through the garbage data later, or analyze the data in advance to make the hard decisions first. Option one seems less expensive and faster, but it’s not. In fact, it can end up being more expensive in the long-term and can take the glean of your fresh new system and reverse it into a weird, morphed version of the legacy system you wanted to get rid of in the first place. So, the best way to start your legacy data migration is to get a team together and set up some parameters for the data. Ask questions such as:
- What’s our criteria for getting rid of unnecessary data? Should we establish a cutoff date? Accounts with no new Opportunities added since 2010 are not migrated? Activities closed prior to 2015 are not migrated?
- What kind of attachments will we use in this new system? Are they all current/viable?
- Are there any outdated fields or criteria we need to get rid of or update to fit the new system?
- How will the configuration of the new system translate with our old data? Do we need to transform any of our existing drop-down values to match?
2. Make the Investment
When it comes to archived data, experimentation can be a dangerous game — the data is too valuable. If your IT team doesn’t have a history with legacy data migration, take the safer route and invest in a migration partner. You’ll increase your probability of success and lower overall project time. Plus, you’ll free up your IT team to work on major projects in the queue instead of investing all their time trying to figure out the right legacy data migration strategy. It all comes down to the cost versus savings of “winging it”.
3. Choose the Right Partner
If you take the advice of tip #2, your next step will be finding an expert to do the job. How do you recognize an expert? Talk to your industry peers, do research on sites like G2Crowd, and find out if the company you’re considering has done other legacy migrations. Ask your IT department to speak directly to a migration partner to address any complicated technical questions. Many companies (StarfishETL included) will give you the option for a free sample migration to help uncover the internal structural issues that may be lurking in the shadows. With that information in hand, you can start to understand the baseline of your data and how much time and energy you can anticipate spending on the migration.
4. Extract the Data First
Any good migration partner will tell you to always extract your data from the legacy system in its raw form and save it to a delimited flat file format. The “raw form” I’m referring to simply means that you pull the data as-is, without any transformations. It’s easier to audit legacy data coming out of the system when you aren’t trying to match it to the new fields right off the bat.
If you were to change the fields and then realize a problem exists, you would have to go back into the legacy system to verify. Using the raw data in the flat file is the easiest way to filter through and audit the information.What’s also helpful about the delimited flat file format is that it gives any technology program access to import the data, and, it gives you the flexibility to work on that data extraction simultaneously as you try to determine how it will be structured in the new system.
It’s extremely rare to be able to simply pull out your data from the legacy system without having to correct some of it. Your migration partner can help you recognize the challenges by looking at your flat file and helping you create a data map to match the old elements to the new ones. Standard maps will map out-of-the-box fields, but clients must tell their migration partners how they want the custom data mapped.
5. Establish Data Transformation Rules
Once the legacy data has been extracted, it’s time to start thinking about transformation rules. What approach should you take to transform the records? How can you determine which format will deliver the best data quality? For example, if your legacy system has a customer record with various subtypes, how will you transform those to make the most sense in the new system? It will depend on your situation, but you can start by asking yourself questions like these to get your data juices flowing:
- Are we defining transactions based on time or event?
- How are we going to handle invalid/null data in the new system?
- What values does the new system come with? How will we translate those?
- What’s our definition for “historical data”? When should data be considered historical?
- Which of these fields are used for reporting?
6. Address Reporting Discrepancies
When you’re wrapped up in data maps and transformation rules, it’s easy to forget about reports. Don’t make this classic mistake! Reports generated in the legacy system may be restricted by the parameters of the new system and may not fully translate over. Your old methods of tracking and reporting may not even exist. Analyze the elements of your legacy system, and understand the reporting functionality of your new system, and determine how you will report on that data in your new system. Document how you need to see the data from a reporting standpoint and make sure your migration partner is aware of it.
Use these 6 tips to stay organized as you begin your legacy data migration journey. By following the right procedures and keeping tabs on how your reports translate, you can avoid roadblocks to your successful legacy data migration. Unlike Forrest Gump, you’ll know what you’re getting and have the structure to turn your data into a powerhouse for your business.
If you’re looking for a migration partner with legacy data migration experience, I hope you’ll keep StarfishETL in mind. G2Crowd didn’t rate us one of Chicago’s Best Business Softwares for nothing! Shameless plug, but seriously, email us at email@example.com if you need help.