There are particular requirements that must be met before data can be imported into Salesforce, as well as best practices that should be followed when either importing manually or scheduling jobs.
Make Sure Data is Clean
When it comes to importing data, you want to make sure your data is clean. For example, are all of your email addresses correctly formatted? It is important to ensure the data you put into your org sets an example for data to follow.
Set up Data Syncs Properly
If jobs are running to sync data from one system to another, make sure methods are in place to prevent bad data from overwriting good data. If there is a bidirectional sync, for example, make sure there is not a cycle of updates that will occur and overwhelm the system.
Run a Test Import
When importing data, start with a small, limited number of records. You want to ensure the import performs as expected, and if not, you have a way to keep it from spiralling out of control. Once you feel comfortable, you can import larger batches of data.
Keep a Backup of Your Data
Perhaps the best practice of all when importing data is to have a backup of your existing data in storage before any data is changed. This is mostly true for manual methods that use Data Loader or Data Import Wizard.
In the event mapping is done incorrectly, or values should not be replaced, reversing the changes can be done quickly if a file of existing data was saved before the change.
Spot Check your Data after Importing
After imports are completed, manually check the records to ensure the data looks correct. For example, you can choose three records from your import and look up those records in Salesforce. Once you’ve located those records, you can check the data to make sure the fields are correct.