Standard CaliberMind Normalization and Deduplication
July 11, 2025
Out of the box, CaliberMind runs some data normalization and deduplication processes to clean up and streamline your records. These include the following:
Field Normalization
When running transforms and standardizations, CaliberMind will:
- strip out special characters
- lowercase emails
- change Vice President to VP, etc.
- fix state and country codes to 2 letter codes to assign regions
- parse out UTM strings with different naming to consistent column names
Field Standardization
We will standardize fields such as:
- Industry value standardization
- Job role, department
- Region / geographic codes / longitude and latitude
Person Deduping
- We keep all duplicates out of the box, but pick a winning record to inherit all the web activity
- Since web activities are based on Email, there could be more than one person record with the email
- CaliberMind will pick the person record that was most recently created to match up the data
- We also have an optional workflow to merge CRM records, which can be previewed before we turn it on
- cm_lead (unconverted or orphan leads) + cm_contact (crm contact) = cm_person all people leads and contacts
- Each orphan tries to match to an account using email domain and website
- Same tie-breaking process is used to pick a winning account among duplicates with the same domain — e.g. the account with the most opportunities or activity
Company Deduping
- We keep all duplicate accounts out of the box, but pick a winning account to inherit all the orphans matching up
- We append website to accounts missing website by looking at related contacts’ email domain
- We have an optional company merge flow that can merge accounts in the CRM, which can be previewed before turning on
Event Standardization / Deduping
- There is native deduping that will not allow the same event twice on the same day
- There is campaign member deduping that does not allow the same person to respond to the same campaign twice
- All other events are kept in the event table and we attempt to match to a CRM person id and CRM company id