Sync data from
Snowflake to Salesforce

Cloud data warehouses like Snowflake have become the standard for analytics in establishing a single source of truth for data teams. However, sending Snowflake data to Salesforce can be challenging and complicated. The good news is we've compiled a list of the many different ways you can sync data from Snowflake to Salesforce.

Option 1: Reverse ETL with Hightouch

For fully automated data syncs from Snowflake to Salesforce, Reverse ETL is the best option because it's not only the easiest to use but it's the least prone to failure. Instead of dealing with manual CSV files, third-party APIs, or custom code, Reverse ETL syncs data directly from your warehouse to your operational tools (e.g., Salesforce.)

Using Hightouch for Reverse ETL, you never have to import your data. Hightouch is a SaaS platform, which means you don't have to install anything locally on your computer. Record limits do not burden the platform, and the first destination is free.

With Hightouch, you can leverage all of your existing Snowflake tables and data models and sync that data directly to over 100 different destinations. You can run your data syncs manually or schedule them on a set cadence that you define (e.g., after your transformation jobs finish running in Snowflake.) 90% of all Hightouch syncs finish within 30 seconds, and this includes millions of records. There’s even a visual audience builder for non-technical users.

Step 1: Connect to Snowflake and Salesforce

You can connect to Snowflake using your standard information, and you can connect to Salesforce through 0Auth.

Step 2: Define Your Data

With Hightouch, you can write standard SQL or leverage your existing tables and dbt models. If you use Looker as your BI tool, you can even query using Looks.

Step 3: Configure Your Settings

Once you've defined your source data, the next step is to determine how you want your data synced to Salesforce. You can choose to upsert, update, or insert your data.

Step 4: Map Your Data

After you've selected your primary key, you can map your data to the appropriate fields in Salesforce.

Step 5: Run Your Sync

You can run your sync manually, on a set interval or custom recurrence, use a cron expression, or even schedule your sync to run after your latest dbt job has finished running.

Option 2: Salesforce Bulk API 2.0

Salesforce has a lot of different APIs, but you'll most likely want to leverage the Bulk API 2.0 if you want to read from Snowflake and write to Salesforce. This is the easiest API to upload, query, or delete records in Salesforce at scale. Salesforce’s documentation has more details on how it works.

Before you can use the Salesforce API, you’ll need to choose a library to connect your source and your destination. Depending on your language choice, this will vary, but JSforce is a great option for Node.js developers.

After you’ve chosen your library, you’ll need to write a script to pull out the records you want to sync to Salesforce, and then you can ingest your data into Salesforce’s Bulk API. Of course, it’s never as easy as just calling an API.

There’s a lot you need to account for, whether it’s authentication, error handling, logging, rate limits, or batching. You also have to worry about various edge cases and parallelizing. If you’re considering building your own in-house pipeline to send data from Snowflake to Salesforce, check out this guide on integrating third-party APIs.

Option 3: Data Import Wizard

Data Import Wizard is a native feature that Salesforce makes available to all editions except Personal and Database.com. This tool lets you upload CSVs directly to Snowflake. You can import data for many standard Salesforce objects like accounts, contacts, leads, and campaign members. You can also import data for custom objects as well.

The Data Import Wizard only handles 50,000 records at a time, so it’s not a dedicated data integration tool. The tool is really only designed to handle CSVs. You have no ability to perform deletes or exports with this tool. All imports have to be done manually since there is no scheduling option.

Data Import Wizard is available in the Salesforce Quick Find bar. Once you’ve chosen selected an object and a primary key to uniquely identify your records, all you have to do is upload your CSV to Salesforce and map the data from your CSV to the appropriate columns in Salesforce.

Option 4: Data Loader

Another option for sending Snowflake data to Salesforce is Data Loader. This feature is free for Enterprise, Unlimited, Performance, Developer, and Database editions of Salesforce. It’s basically just a more robust version of the Data Import wizard, and it supports up to 5,000,000 records per import and can be used for both importing and exporting.

Data Loader is a client-based application that you have to run locally on your computer. There’s a command-line interface that’s available for Windows users to give you more options around configurability.

You can find the installation files Data Loader in the Salesforce Quickfind bar. Getting started with the Data Loader is a bit more complicated than the Import Wizard because you need a Java Development Kit (JDK) to run it locally on your computer.

Without changing anything, the maximum number of records that Data Loader can process in a single insert, update, upsert, or delete operation is 200, so you’ll need to configure the tool to use the Salesforce Bulk API.

The Data Loader interface isn’t that intuitive, and the UI feels like it’s been around since the early 90s, so it can be a bit complicated to import your CSV and map your data fields to the appropriate object in Salesforce.

Option 5: dataloader.io

Mulsesoft originally developed dataloader.io before Salesforce acquired the company back in 2018. It offers basically all of the same functionality as the Data Loader, but it’s available as a SaaS offering you can access through your web browser, so it’s much easier to use than Data Loader.

The free version of dataloader.io only supports 10,000 records per month, so if you need to import more records, you’ll likely need to purchase an actual license. Licenses start at $99 a month per user for the professional plan and $299 for the enterprise plan. The professional plan gives you 100,000 records, and the enterprise plan gives you unlimited.

Much like the previous options, Dataloader.io has a UI where you can upload CSVs and map your data to the appropriate objects in Salesforce. In order to schedule imports with dataloader.io, you have to save an updated CSV with the exact same file name as your previous import.

Option 6: CRM Analytics Snowflake Connector

CRM Analytics is a native Salesforce product that provides a native connector for Snowflake where you can sync data to and from Snowflake. The product starts at $125 per monthly user if you want to access the data connectors. Connectors themselves can sync up to 100 million rows or 50 GB per object.

CRM offers three main connection types:

• Input connectors sync data directly to Salesforce.
• Output connectors move data out of Salesforce.
• Live connectors create a live connection to your source and query data directly in Salesforce.

In order to use CRM analytics first have to build recipes and workflows to map your data to the appropriate fields in Salesforce, and there’s no easy way to update, insert, or upsert fields. Unfortunately, the live connector is only practical for reporting since you can only build charts, dashboards, or tables with the data.