Salesforce and How to Avoid Dirty Data

Previously, we’ve discussed using Salesforce for data collection and also defined dirty data and its costs. Diagnosing how data gets dirty is a great starting point in setting up systems and preventive measures. The usual suspects for dirty data come from humans and  missing or faulty processes.

Human Error

A major contributor to dirty data is human error. Humans are bound to make mistakes even when they are highly skilled and motivated. And of course even the best AI cannot replace a human so what can be done? Using some of these strategies can help counteract the inevitable.

How to Keep Data From Getting Dirty

Automated processes and validation rules hold significant value in time savings and data hygiene. However, these tools become less effective when they either don’t exist or haven’t been kept current.

Are there consistent human errors like a field being updated incorrectly? If so, it’d be wise to consider implementing an automated process to perform that task. For example, this could be addressed by automatically populating a date field with the current date or calculating a value based on other data already known rather than manually inputting the information.

Are key fields being left blank? In these cases, making those fields required or creating a validation rule could be the best solution.

In a multi-step process, are steps being forgotten? If so, creating a workflow to walk users through the process and including automation and validation rules where needed, could be the answer.

When tracking a new value for a set of records, consider the impact for all existing records. For example, adding a field titled, First Contact Source to a contact record to start tracking how individuals first learned of the business—then creating simple reports off of this so users have a predefined list of options from which to choose. This is really important information so the field is required. 

Seems straightforward but what happens to all the existing contact records? Setting a default risks people not updating to the correct choice causing data to be inaccurate (dirty).

Leaving it blank isn’t the solution. When someone comes in to edit a different value on this record, they’ll get an error message as this field is required. The person performing the edit may not know how this client came to the business so what do they do? In haste, they may choose to just pick any value from the list. If their guess isn’t correct, and most likely it won’t be, that data will be inaccurate and dirty. They may choose to go dig for the correct information causing a delay in getting their intended update completed. Or worse, they may forget to make the update altogether. Here again, inaccurate date equals dirty data

In these cases, we advise a multi-step approach:

• Create a value in the list of choices to indicate this is a legacy record and the info may not be known.
• Automatically update all existing records to this legacy record choice.
• If possible, consider hiding the legacy record choice as a valid option going forward.  

When is Data Clean Enough?

Data cleaning is very much like house cleaning when you have children. Shortly after completing the job, the house starts to get messy again—very quickly. As long as users are interacting with data or adding new records, it will inevitably lead to dirty data. The best course of action is to craft a plan for data cleansing. Whether that task is on an annual or semi-annual basis will depend on the amount of data and staffing resources.

Let’s Get to Cleaning

A CRM, such as Salesforce, is generally the key place to start when considering data hygiene—whether looking at data structure, duplicate records, or implementing processes to prevent data from getting dirty. Phase2 is here to diagnose the cost of dirty data, guide in cleaning the dirty data, and help prevent future data dirtiness. 

We hope you enjoyed reading the Phase2 blog! Please subscribe below for regular updates and industry insights.