Best practice is to set up data quality dashboards showing data quality scores by user and provide some incentives for users with high data quality scores like the opportunity to win an iPod. For example, one Salesforce.com customer introduced a Data Quality dashboard in which records owned by individuals follow a simple score. For example, all records are rated between 0-100. So if there are five attributes on a record, then each attribute is worth 20 points and this tells us whether the record is complete.  The net effect is that each record can instantly tell you how qualitative it is and you can then draw together stats for the user.  Click on the graphic for an example Dashboard for the data quality scores.

In this example we rate Account, Contact and Opportunity details but you could rate anything. The attributes that we measure are listed below each of the first charts. By measuring the quality of the data entry activity rated by user, you can spot persistent offenders and reward good behavior.

Here are example screens shots and custom formula fields that should work against any salesforce.com system as they are based on standard fields.  Even if you have extended your Salesforce system, these will work.  You just need to review what attributes are important to you and adapt the formula accordingly. Customers looking to implement this should also consider adding data quality scoring on other objects like Leads and Cases.

Account - Data Quality Score

IF( ISPICKVAL( {!Industry} ,""), 0,20) + IF( ISPICKVAL({!Rating},""), 0,20) + IF( LEN({!BillingCity}) = 0, 0,20) + IF( LEN({!Phone}) = 0, 0,20) + IF( ISPICKVAL( {!Type} ,""), 0,20)

Account - Data Quality Description

IF( {!Data_Quality_Score__c} =100,"All Account Details Captured", "Missing: "&IF( ISPICKVAL( {!Industry} ,""),"Industry, ","")&""&IF(ISPICKVAL({!Rating},""), "Rating, ","")&""&IF( LEN({!BillingCity}) = 0, "Complete Address, ","")&""&IF( LEN({!Phone}) = 0, "Account Phone, ","")&""&IF( ISPICKVAL({!Type},""), "Type",""))

Contact: Data Quality Score

IF( LEN( {!Phone} ) = 0, 0,20) + IF( LEN( {!Email}) = 0, 0,20) + IF( LEN( {!MailingStreet} ) = 0, 0,20) + IF( LEN({!Title}) = 0, 0,20) + IF( ISPICKVAL( {!Salutation},""), 0,20)

Contact: Data Quality Description
IF( {!Data_Quality_Score__c} =100,"All Contact Details Captured", "Missing: "&IF( LEN( {!Phone} ) = 0, "Phone, ","")&""&IF(LEN( {!Email}) = 0, "Email, ","")&""&IF( LEN( {!MailingStreet} ) = 0, "Complete Address, ","")&""&IF( LEN({!Title}) = 0, "Job Title, ","")&""&IF( ISPICKVAL( {!Salutation},""), "Salutation",""))

Opportunity: Data Quality Score
IF( ISPICKVAL({!Type},""), 0,20) + IF( LEN( TEXT(MONTH({!CloseDate}))) = 0, 0,20) + IF( {!Amount} <= 0, 0,20) + IF( ISPICKVAL( {!LeadSource},""), 0,20) + IF( LEN({!NextStep})=0, 0,20)

Opportunity: Data Quality Description
IF( {!Data_Quality_Score__c} =100,"All Opportunity Details Captured", "Missing: "&IF( ISPICKVAL( {!Type} ,""),"Type, ","")&""&IF(LEN( TEXT(MONTH({!CloseDate}))) = 0, "Close Date, ","")&""&IF( {!Amount} <= 0, "Amount, ","")&""&IF( ISPICKVAL( {!LeadSource},""), "Lead Source, ","")&""&IF( LEN({!NextStep})=0, "Next Steps",""))

To take this a step further, in theory what is it that an organization wants to do once their data is clean?   Maybe you want to use an account scoring process to aid your sales team in identifying who they should be targeting based on criteria such as market size, industry sector, number of employees, revenues, interests, etc. This approach would work with the data quality measures discussed and take it further by providing insight into how an organization can then use clean prospect/customer information to work out who to target.  For example, in the account scoring process, you can define a series of key account criteria and then determine a formula (using weighted criteria) to derive an account score. The formula would adapt according to the status of the account, i.e. adapt once the account converts from a prospect to customer. The account scoring would provide trending functionality based on the present and historic scores to determine a rising, static or falling value. The solution would deliver sufficient granularity in its calculations (rather than just a final numeric value against the account) that can then be used to mine the data to identify more trends. We recently saw a demo of how one organization had used this account scoring approach within Salesforce. It was quite compelling. 

Have a CRM question? Askwendy@crmsuccess.com