Salesforce Data Quality: The Elephant in the Room
Salesforce Best Practices, Salesforce Data

Salesforce Data Quality: The Elephant in the Room

It’s no secret that organizations struggle to keep their CRM system data up-to-date and clean. With daily website inquiries, years of legacy customer data, manually entered contacts, integrated marketing systems, purchased lists, customer support requests, and the inevitable integration to the finance system; there are just so many opportunities for data quality to fall apart.

While the sources of data quality issues are varied and numerous, the steps to resolve them are relatively straightforward. When followed, these best-practices can bring data quality improvement to even the messiest of systems.


The first step has to be defining what good data means for your Salesforce system. Every instance of Salesforce has subtly different goals, therefore no generic definition will do. Creating a Data Quality Management document that contains these agreed upon definitions and grows and evolves with your Salesforce system is essential from the start.

Perhaps this sounds obvious, but its a step often skipped. For many, the idea seems like a waste of time as everyone knows what clean data looks like. System Administrators will often jump right in and begin to purge incomplete records and merge duplicate accounts, only to find themselves repeatedly going back to the business asking questions like, “Is there value in keeping a contact with a valid email address, but no last name?”

It’s important to start with the end in mind. Below are questions to consider when developing your Data Quality Management document.

  • What is the bare minimum data required for each record for it to still be useful?
  • Are there ever situations where duplicate records are allowed/necessary? Why is this the case?
  • How will things such as  LLC, Inc, Corporation be abbreviated in the system?
  • Is Salesforce integrated with other systems? How do those systems define clean data? Do the definitions align?


The next step in improving data quality is to determine if the creation of dirty data is ongoing. You can’t make real progress if while you are cleaning existing data, the system is busy creating more dirt. You need to draw a line in the sand. Stop the creation of new dirty data today, and then at a minimum, you will know that all new data going forward is at a level of quality that’s acceptable.

To actually do this, start by looking at every point of data creation in your system. Ensure that each point is only allowing new records to be created that align with the defined quality standards.

Below are some common problem areas:

  • Website Inquiries – Are these coming in as Leads and is the website requiring enough information to be entered, so they are useful?
  • User-Created Records – Does the system have the appropriate required fields set for each object? Are users set up with the correct level of security to see if they are creating a duplicate record? Are there additional Validation Rules needed to ensure the quality of newly created data?
  • Integration Created Records – Does the integrated system provide the minimum threshold of data when creating a new record? Is it clear which system is the master of each integrated field? Is it clear how to clean up a data issue in the integrated system if that data impacts Salesforce?
  • System Generated Records – Are there any Workflow rules, Triggers, Flows or Process Builder automation that are creating new data records. Do those records meet the minimum threshold for data quality? Are all custom fields on the Lead object mapped to a field on either Account, Contact or Opportunity field for Lead conversion?


With Salesforce no longer generating new records with questionable data, now is the perfect time to clean your existing data. This can be done easily with a data cleansing tool such as DemandTools (which we highly recommend) or if you are working with smaller data sets, the data extracted from Salesforce using the Salesforce Data Loader and the standard Salesforce merging functionality in both Salesforce Classic Merge Functionality and the new Salesforce Lightning Experience Merge Functionality. may be more than enough.

No matter how it’s accomplished, it’s essential that Account issues like ACME inc. vs. ACME Incorporated or Contact issues like Mr. John Doe vs. Dr. Jonathan Doe are resolved fully. Easier said than done; I know.


Unfortunately, even after successfully completing all of the steps above, the work of data quality management is never entirely done. Not every issue can be resolved with automated steps, and there are limits to how many validation rules you can put in place before your end users scream.

To maintain truly exceptional Salesforce data its essential to put in place some form of the following steps.

  • Data Creation Monitoring – Even if it’s only on a monthly basis, just looking at a random sample of newly created records, gives you the opportunity to catch issues before they go too far.
  • Train the End Users – It is a best practice when training new users of Salesforce to include some aspect of data quality training. Highlighting what information is required to create a new record gives them the opportunity to collect that information when they interact with their clients.
  • Publishing Data Quality Statistics – Creating metrics around data quality and sharing the results over time can have a surprisingly positive impact. Dashboards such as the Salesforce Data Quality Analysis Dashboard can be a great place to start. If it’s not measured, the progress you have just made will never last.


If your organization has invested in Salesforce and you believe you could be getting more value from it, it is often a good idea to bring in a professional to have a look. Salesforce can be a good place to start, or you can call a Salesforce consulting services provider like CloudMasonry.