Salesforce Sales Cloud and Snowflake: data synchronisation via Data Cloud

The question of synchronisation between Salesforce Sales Cloud and Snowflake has become central in modern data architectures. With the emergence of Salesforce Data Cloud (formerly CDP) and the strengthened partnership between Salesforce and Snowflake, the options have evolved considerably in 2025-2026. Here's a comprehensive overview.

The context: why synchronise Salesforce and Snowflake?

Companies using Salesforce Sales Cloud generally have two complementary needs:

• Enriching Salesforce from Snowflake, propensity scores, behavioural data, purchase history, external data, so sales reps have a 360° view in their CRM.
• Exporting to Snowflake, CRM data (opportunities, accounts, activities) to analyse alongside other sources in the data warehouse.

Until 2023, these synchronisations were mainly done via ETL/ELT (Fivetran, Talend, Informatica) or custom Apex integrations. Data Cloud changes the game.

The 3 synchronisation patterns

Pattern 1 : Data Cloud as central hub

This is the native pattern recommended by Salesforce. Data Cloud ingests data from Sales Cloud, unifies it around a single customer profile, then shares it with Snowflake via Snowflake Data Share (zero-copy).

• No data movement, data stays in Snowflake, Data Cloud accesses it directly.
• Near real-time latency for critical data.
• Cost: requires a Data Cloud licence (not included in Sales Cloud).
• Ideal for: profile unification, advanced segmentation, marketing activation.

Pattern 2 : Classic ETL/ELT

Still relevant for simple cases or when Data Cloud is not available. Fivetran, Airbyte, or Talend extract Salesforce data via the REST/Bulk API and load it into Snowflake.

• Simpler to set up than Data Cloud.
• Batch synchronisation (hourly or nightly depending on the connector).
• No native near real-time.
• Ideal for: BI reporting, historical analysis, no latency constraints.

Pattern 3 : Custom integration via Apex + Snowflake API

Legacy pattern, still present on older projects. An Apex Trigger or batch job calls the Snowflake REST API to push data.

• Complex to maintain, Salesforce governor limits to respect.
• Difficult to scale on large volumes.
• Avoid for new projects, prefer Data Cloud or ETL.

Focus: Salesforce Data Cloud : what you need to know

Data Cloud (formerly Customer Data Platform / CDP) is the centrepiece of Salesforce's data strategy. It's not a simple connector, it's a customer data management platform that:

• Ingests data from multiple sources (Sales Cloud, Marketing Cloud, web, apps, Snowflake…)
• Resolves identities, the same customer recognised across multiple channels
• Calculates segments and scores in real time
• Activates this data in Sales Cloud (enriched fields, segments available in Flows)
• Shares data with Snowflake zero-copy via Snowflake Data Share

Pattern comparison

Concrete use cases

Enriching Salesforce accounts from Snowflake

Your data warehouse contains data your sales reps don't have in Salesforce: financial health scores, e-commerce purchase history, web engagement. Via Data Cloud, this data can enrich Account and Contact records in Sales Cloud in near real-time, without custom development.

Advanced commercial segmentation

Data Cloud calculates segments from combined Salesforce + Snowflake data. These segments are available in Sales Cloud to filter lists, trigger Flows or prioritise commercial actions.

Unified reporting in Snowflake

All CRM data (pipeline, activities, conversions) is available in Snowflake via Data Share. Your BI teams can join it with other sources without additional ETL and without impacting Sales Cloud performance.

Watch points

• Data Cloud licence, not included in Sales Cloud. Cost can be significant for an SME. Assess whether ROI justifies the investment or whether a classic ETL is sufficient.
• Data governance, with Data Cloud, you unify data from multiple sources. Consent and GDPR questions must be addressed upfront.
• Required skills, setting up Data Cloud requires specific expertise (Data Cloud Consultant, Snowflake). Plan for appropriate support.
• Snowflake Data Share volume, zero-copy has limits. Anticipate volumes and test performance before Go-live.

Conclusion

The choice between Data Cloud, ETL and custom integration depends primarily on three factors: your budget, your latency constraints and your need for profile unification. For companies that already have Snowflake and need near real-time synchronisation with bidirectional enrichment, Data Cloud is the most coherent solution with the Salesforce roadmap. For simple reporting needs without latency constraints, an ETL like Fivetran remains a solid and less expensive option.