Data Loading Process for Sales Cloud Implementation
What are the Key Components of a Data Loading Plan?
Receiving Sample from the Client:
One of the first steps in the process is to get a sample of the client data. This includes all datasets that will be part of the final data migration. The preferred format of this data is .CSV format.
Verifying the Quality of the Sample Data:
Upon receipt of the data, SMBHD will do a quality check on the data validating the following:
- CLARITY OF DATA: Does the data have clear field header labels? Are they easily mapped? This gives SMBHD an opportunity to identify which fields in the data can be mapped to standard fields, and which fields need to be custom created on the client-specific Salesforce objects (accounts, contacts, leads, opportunities, etc.). Once created, SMBHD can easily map all of the fields for a clean data load.
- COMPLETENESS OF DATA: Is every required field for every record within the sample data populated? This is a big issue for the client, because if there is missing data in these fields, these records will not get loaded during the final data migration. Knowing that there are issues with the completeness of the data greatly benefits the client. This is because it gives them time to scrub their data early in the project, and have it ready for a clean data load during final data migration into production. Without this step early in the project, the client could be scrambling last minute to get their project over the finish line.
- CONSISTENCY OF DATA: Does the data have consistent formatting? Specifically, does the data contain consistent naming conventions within each field? It is not uncommon to see clients have poor data governance over their legacy data. For example, a typical issue is with the Account Name. Often, we will see it entered in several different formats across disparate databases, using an account name like “Acme” in one database, “The Acme Corp.” in a second database, and “The Acme Corporation” in a third database. It is best practice to standardize the naming convention for a critical field like account name across all data. The client needs to choose one naming convention and use it across all records in Salesforce.
- RECORD VOLUME: The last thing we do is a sanity check on record count. We primarily focus on two questions, 1) Did the client send the expected number of records and 2) Is the client attempting to load more records that was agreed to in the Statement of Work (SOW)? This QC ensure that all we received all the records expected from the client and that contractual agreement on record count is not exceeded.
Loading Sample Records and Client Review:
Once we are complete with the QC’s, we’ll go ahead and load 5-10 test records for the client. The client will then QC the data to ensure the expected data was loaded and is presenting as expected in Salesforce. One of the primary goals in this step is for the client to have real data in the system for all the mid-project demos and for any pre-UAT testing the client would like to do in the system.
Manual Data Migration (Post-Deployment):
Once the solution has been deployed to production, we will go ahead and immediately upload all the actual production data. Typically, this is Accounts, Contacts, Leads, Opportunities, and any custom object data.
This is the very last step, where SMB will do a final QC on the data. There are two primary QC’s completed in this final stage. One, to validate the number of records loaded matches the number of records in the source data; this is done by object. Then we’ll perform a sanity QC on the data which includes checking 5-10 records having a full review of their field level data: comparing it the actual records in the source data. This is done by object.
There you have it, a simple high-level review of the data loading process for most new projects. We hope this helps clear up any of the big questions you may have about the Data Loading Process for Sales Cloud Implementation. If not, we would be happy to discuss it further so feel free to reach out to us any time.
-Steven B. Carter, Project Manager