14 Data Migration Best Practices for Successful ERP Projects

At To-Increase, we help many customers in their data migration projects. We use and optimize our best practices to ensure the best possible outcomes for the deployment of industry-specific solutions and ERP systems. We also incorporate our findings in our software development and tutorials. Let’s take a look at how we use the Microsoft Dynamics Sure Step methodology in performing successful data migration projects.

Most of our data migration projects involve companies in multiple countries and different time zones. Many of the companies have different business models and lines of business, such as production and sales. Most of the time, the IT team is centralized and experienced in ERP systems and data migrations. That means we learn together and can share expertise.

In our application of the Sure Step methodology, the main stages are:

• Analyze
• Design
• Deploy
• Operation

By following and applying these best practices, you can achieve a successful project outcome.

Analyze

• It’s extremely important to define thescope of the project, and to define from which applications you want to migrate. By involving the key users and stakeholders of the project, you can make sure that you have firm commitments and that the right people are available to verify and approve the results.
• By describing the requirements you can define what you want to migrate and the scope of the data. For example, you don’t want all the data that are like “zzzzz” or “99999” and data that was not used the last two years.
• Test documentation helps you do repeat tests over and over again. When you create the tests by using the task recorder, you know how to create and load the data manually. This will ensure that you can load the data and can reproduce it manually. That will make it also much easier for the migration consultants to test how and if the migration works.
• You can use our team setup to define who is working on the data migration setup to get central tracking and tracing. That makes it easy to assign activities and to make sure that there is follow up.

Design

• Make sure that, before you configure or develop anything, you have the mapping defined and discussed with the business users. Microsoft Excel is a very good tool to help define the mapping. You can store or link the Excel spreadsheets to the requirements you created.

• Keep the data model as simple as possible in every migration. That makes it easier to trace errors and to understand the script, but also to run an update. For example, when some item units are forgotten, you can easily run an update for units to correct.
• When you use files, you should use one folder with a unique file name. For every migration, you can specify a filter for the file names.

Deploy

• When you migrate company by company, you can create migration templates. Think about templates for sales companies and production companies with different migrations and different currencies, or with tax rules which are mapped by using constant values. It’s quite easy to copy a data migration project and use it as a template.

• One common practice is to have a standalone data migration application such as To-Increase Data Migration Studio, in which you can design and develop your migrations. In this application you can restore databases and test data migration projects. The connection can be set up for each type of environment, because you don’t want to load test data in your production company.
• Use one project definition with one set of files for a data migration project. When you run one project, you have to make sure that all the files are available in the right place. The easiest way to verify this is to use one consolidated folder for each project.

Operation

When you export, for example, vendors yesterday and items today, you could suddenly have an item with a new preferred vendor which does not exist in the vendor file.

• Make sure that you use a stable set of data. Some of our customers use a data warehousing tool to create a stable data set. However, a set of files is also very useful. The important consideration is that all data is exported at the same time.
• When a data migration project is executed, you should set up dependencies by using the task setup. This will make sure that you create your vendors first, before you can use one of them as a preferred supplier on your item.
• Performance is one of the key points for a data migration project. We obtain very satisfactory performance levels by using the dependencies on the task level, because everything with the same dependency can run in parallel. In addition, you can split the input which can run in parallel. For more detail see our earlier blog post Fast Data Migration to Microsoft Dynamics AX 2012.
• During the import we validate all the data and apply the business rules. This gives you the same result as keying the data in! But it makes the migration reliable, so you can trust your data. The project history shows the results by using the validation. From there, you can correct data and reprocess it. And, of course, you should use your defined tests to verify the result. Test your setup several times. With this fast performance, you can run a full data migration several times a day.

If you have any questions regarding this blog post or To-Increase Data Migration Studio, please get in touch with me.