Define Data Quality
First, decide what you mean by "bad data". Get input from Sales, Operations, and other teams that regularly use Salesforce. You should also define the qualities of "good data." Consider data accuracy and completeness and define what constitutes an appropriate record for Marketing to generate and Sales to pursue.
Determine Key Fields for Your Salesforce Organization
Together, your teams can determine a set of key fields for the Salesforce records you use. Key fields can collect information critical to business strategy and operations, and you may decide to make them required. To decide which fields are key to your marketing and sales processes, understand the different stages a lead will go through at every point in the funnel and the possible paths and outcomes.
Some factors to consider are your team’s lead status definitions, especially with respect to sales outreach approach, the sales hand-off process, and nurture and re-nurture practices.
Identify Bad Data
Data quality problems include missing data, inaccurate data, nonstandard formatting, and old or low-quality leads. One way to start identifying bad data is to run Salesforce reports of inactive leads. The activity period will depend on your company, the last time you performed an audit, and the size of your sales team.
For example, run a report on the Lead object with the criteria IF don’t have activity in previous 6 months AND are in Pre-MQL status. When leads are inactive for long periods, the data on the record may be inaccurate or incomplete for a sales rep take action on. However, there may be reasons to keep stale Pre-MQL leads, so make sure to get alignment with your marketing team before including this criteria for cleanup.
If you have an old lead list or a list from a source known to be low quality, you may want to look into removing these records, as well.
Back up Your Data
Make sure you back up data before making any changes and have a recurring back-ups scheduled for your databases. This way, if you need to revert a change, you can.
Append Missing Data
Using Data Append in Radius, you can append fields to your Salesforce records in three ways.
- Select Radius fields to append to new records, and we’ll fill them each time you deploy new accounts, contacts, and leads to Salesforce.
- Append data to all your historical Salesforce account, contact, and lead records in a bulk, one-time update.
- Append Radius fields to future account, contact, and lead records you add to Salesforce.
For more information about our data append solution, check out our help: Data Append.
Remove Bad Data and Duplicates
Whenever you have duplicate data within multiple systems, it’s easy for information to get out of sync. For each business process, determine which system should be the "master," and integrate your systems so changes in one system automatically update the others, creating a clean system of record.
Radius' Data Stewardship service can run an analysis to identify all of your outdated or duplicate records and help you remove them.
Maintain Data Quality
Data cleansing is not a one-time project. Clean data requires frequent and systematic maintenance.
Simplify Salesforce Data Entry
Make it easy for your sales reps to enter and maintain clean data. Keep Salesforce record page layouts, validation rules, and record types as simple as possible, and communicate changes and requirements clearly. Make sure everyone entering or updating record data is aware of required fields and formatting. Salesforce maintains extensive documentation on these kinds of administrator processes and tools.
Set Up Standardization Processes and Tools
If you haven't already, implement validation rules to increase the quality of incoming CRM data. Validations rules check your data for specific criteria before it is saved and help make sure that Salesforce users enter the correct information. You should also use tools for formatting and validating fields that require a specific format, such as addresses. For example, use State and Country picklists in Salesforce to make sure all address data is uniform.
Continually complete missing data by running analysis on fill rates and data accuracy. Set up a frequent cadence to perform analysis on fill rates across key fields and leads with no activity.