Every growing business will undertake a data migration at some point. While it is both an essential (and even daunting) task, the health of your data is critical to business growth and success. Fortunately, there are measures you can take to ensure a successful migration project. We’ve summarized them here in six steps toward a practical test plan.
Understand the Data Your Business Runs On
Modern business runs on data. It is the fabric that keeps companies connected, informed, and competitive. Before attempting a data migration, you must understand what data you’re collecting and how it drives your core business processes.
If you are migrating from a legacy CRM system, this step is simple. You’ve already put considerable thought into your business processes and all your data stored in a single location.
But if you are trying out CRM software for the first time, you will need to go on a data finding mission. Look for data siloed on an individual’s computer, a department’s server, or a third-party application, or even on business cards or post-its. You’ll need to find your company data in all its hiding places and all its formats.
Once you gather all the data, you’ll need to organize it. The people who use the data daily are your unofficial experts. They will best understand the relationships between the data and the business’ core processes.
After you’ve cataloged the data, you can define your goals for the target system. What problems are you hoping to solve with your new software? If you are switching CRMs, make a note of any inefficiencies in the legacy system.
Before You Start: Create a Data Migration Timeline
Once you understand the size and scope of your data, you can accurately plan the data timeline. Downtime is one of the hidden costs of data migration. It can result in lost customers for 37% of businesses or 17% lost revenue. Having a timeline with dependencies and milestones can minimize the risk and ensures the migration plan goes as smoothly as possible.
Six Steps to an Effective Data Migration Test
While every business is different, the following six steps are relatively standard for most data migration projects. When assigning a timeframe for each step, consider the following: size and complexity of data, migration method, number of key players, and optimum delivery time (as determined by management).
1. Data Cleansing
The more data you have, the more costly and time-consuming a data migration will be. It makes sense to examine your data and limit the migration to what is essential. Clean up your data by:
- Removing historical data you no longer use;
- Deleting duplicate records;
- Filling out incomplete records.
It is also a good idea to evaluate your data structure. Remove any unnecessary objects, fields, tags, and other entities. Try to ensure your target system doesn’t inherit the problems and inefficiencies from the old one.
During this process, you should carefully document your data’s scope and what you will include and exclude from the migration.
2. Choose Your Data Migration Method
Many CRMs have a native migration tool. First, you must save your data to CSV (comma-separated values) files. CSV allows you to retain the structure of your data in tables. Then you upload the files to your target CRM. A CSV migration is one of the easiest and least technical methods if the new and legacy CRM systems are similar.
Some CRMs do not accept all types of data via CSV files. For example, the popular CRM, HubSpot, cannot import notes, calls, tasks, or emails via CSV. So, there’s often a trade-off between simplicity and scope. You can learn more about CSV file migration here.
API endpoints are another method for migrating data. There are usually no restrictions on where data locates itself. It would help if you had a team of data specialists to perform an API migration.
A third option is to use a third-party service partner or app. For most businesses, this is the most cost-efficient and time-saving method for data migration.
The timing of your migration approach is another important consideration. There are two data migration schedules: big-bang and trickle.
- Big-Bang Migration: The entirety of your database transfers within a set period. The downside to this method is system downtime. The upside is you get the entire migration done in one go.
- Trickle Migration: Trickle data migration is completed in phases while the legacy and target systems run parallel. This avoids downtime and its associated costs. Trickle migration is also reliable for continuous migration.
3. Backup Data Before Migration
Whatever migration method you use, you must backup your data. Data migration is a huge undertaking. An estimated 83% of migrations fail or run over budget. Having a backup ensures you won’t lose the data your company runs on.
4. Examine the Source and Target CRM Systems
You’ll need to plan how the source data fits into the new CRM system. You should have documentation from the data cleansing step, which tells you the data types, structure, and the number of records migrated.
Then examine the target system. You will need to know how each field and entity maps onto the new system. If you have any custom requirements, build them out in the target CRM before the sample migration.
From there, you can create a template for how the data maps from one system to another. For example, if there is a significant disparity between CRMs, you may prefer to use a third-party service provider. Alternatively, you can use a data migration app. Most of these apps offer data matching templates.
Trujay’s migration app is unique in its flexibility. It automatically maps the data while still allowing you to customize fields and other entities.
5. Run a Sample Migration
A sample migration is when a subset of data is migrated. This dry-run approach allows you to catch mistakes before migrating your entire database. It’s also much more efficient than migrating your complete dataset repeatedly as you perfect your mapping. Ten percent of your records is a typical amount.
Running a trial migration can also help you prepare for any system downtime. Once you know how long it takes to migrate 10% of your database, you can extrapolate the entire migration timeframe. If it exceeds your project timeline significantly, you can try another migration method or tool.
After running a sample migration, it’s advisable to compare several records in the target and legacy systems. Then, IT can check the records against the mapping document for accuracy. Meanwhile, CRM users can test the new system for usability and intent. Team members know the data they use intimately and can screen issues IT might not notice. They can also better evaluate whether the new CRM is meeting its stated goals for the business.
If you are using a data migration tool, they usually come with sandbox testing. This means you can attempt multiple trial migrations. In addition, these automated tools make it easy to tweak the mapping if you get any unexpected results, and the process takes mere minutes.
Manual users will need to trace the original dataset, determine where each errant data point came from, and then recalculate the field values. This method is time-consuming and more prone to human error.
6. Complete Your Data Migration
Once you’ve completed a trial migration that looks good to both IT and system users, you can complete the migration process. Again, having a plan, timeline, backup, and trial run will prepare you for successful data migration.