Census

Census is a reverse ETL (Extract, Transform, Load) platform that enables data teams to sync data from their data warehouse into CRM systems, marketing platforms, customer success tools, and other operational business applications. By making the data warehouse the authoritative source of truth for customer data and business logic, Census bridges the gap between the analytical data stack and the operational tools that sales, marketing, and success teams use daily.

The reverse ETL model that Census implements is the inverse of traditional ETL. Traditional ETL moves data from operational systems into the warehouse for analysis. Reverse ETL moves analytical insights and customer data from the warehouse back into operational tools, enabling those tools to benefit from the rich, curated data that exists in the warehouse but is otherwise inaccessible to sales reps, marketers, and customer success managers.

Models in Census define which data to sync using SQL queries or dbt model references. A model selects rows from the warehouse with the fields that should be synced to the destination system. This SQL-based approach leverages existing warehouse expertise and allows the full power of SQL transformations to prepare data for each specific destination's requirements.

Destinations in Census cover the most widely used business applications including Salesforce, HubSpot, Marketo, Braze, Iterable, Klaviyo, Intercom, Zendesk, Pipedrive, Freshsales, Amplitude, Mixpanel, Google Ads, Facebook Ads, LinkedIn Ads, Slack, and many others. Each destination has a native integration that understands the data model of the target system and handles upserts, associations, and relationship management automatically.

Syncs define the mapping between the columns in the Census model and the fields in the destination application. Users configure which source column maps to which destination field, how records should be matched (using primary keys, email addresses, or other identifiers), and whether the sync should create new records, update existing records, or both. Syncs run on a defined schedule, keeping the destination application current with the warehouse data.

The incremental sync architecture in Census is designed for efficiency and reliability. Rather than re-syncing all records on every run, Census tracks which records have changed since the last sync and only processes the delta. This approach minimizes the number of API calls to destination systems, reducing costs, avoiding rate limits, and keeping sync times short.

Audience hub in Census enables creating dynamic audience segments based on warehouse data and activating them across advertising and marketing platforms. Segments are defined with SQL or a visual segment builder, and Census handles the continuous synchronization of segment membership as warehouse data changes.

Census is compatible with all major data warehouses including Snowflake, BigQuery, Redshift, Databricks, PostgreSQL, and others, fitting naturally into existing modern data stack architectures.