If you plan on cleaning Salesforce you should ensure that the following tasks are done prior to your data cleansing project.
Create a data quality dashboard: You can create a data quality dashboard that will enable you to see the reps that are adding incomplete data, source of bad data.
Create alerts for massive data uploads: Create an alert every-time someone add many leads or contacts to Salesforce. This will maintain the integrity of your CRM.
Formulate policies and create strategic guardrails for data: There is a need for users to comprehend the entire process and policies which surround well-groomed data. Provide a data holding room and help them understand the importance of quality checks – running leads through marketing programs such as Eloqua, Pardot, Marketo or any other before dumping them into the system.
Validation and monitoring: Data should be monitored to ensure that all required fields and information are correctly filled. If there is missing data, use third party tools such as data.com, Hoovers, Zoominfo and other data providers to append missing data.
Consistent and Standardize fields: Certain fields should be normalized and standardize so that you can easily segment, route leads and run reporting. An easy way of making data consistent is to turn the field into a drop-down this will not allow anyone to type in free text and your data will be identical. If you are upload lists manually, make sure to normalize your data prior to upload so it matches your drop down values.
Once you have established data governance, you are now ready for cleansing your Salesforce. Here are a few things you should consider:
Eliminate redundant duplicates: It’s a colossal waste of human effort and funds going through duplicate data. A more efficient way is to use a de-duplication tool or hire a data cleaning company to do the de-duping for you. Be careful when de-duping as you do not want to lose valuable information.
Normalizing Historical Data: Historical data should be normalized after you have implemented value selection. To clean it you can use a look up table and run a simple vlookup command in excel if you have a few values or use more advanced fuzzy logic techniques.
Categorizing Non Current Data: Most marketers today use lead nurturing, but few companies do the same for account and contact data. If you are using status you can identify which accounts are currently active in your sales status, by looking at last activity date, date create and other date fields you can segment the data even further. Having a good process on how to deal with accounts that are no longer part of your sales cycle will go a long way to ensure high data integrity.
Data Appending: If a lot of your records are missing data that is critical to your decision making, sales or marketing you can use third party data appending providers. While, this comes at a highest cost you can minimize it by only focusing on key accounts or accounts that are recent. Another solution is to engage your sales and marketing teams to help append that data using smart forms in your marketing automation tool or asking sales reps to ask certain questions when they are talking to a lead or contact.
If you are serious about getting your salesforce cleaned you can contact a data cleaning company such as StrategicDB to have your data de-duped, cleaned and normalized.