Best ETL Tools for Salesforce (2026): Top 9 Compared

dataloader.io is a web-based tool for importing, exporting and deleting Salesforce records, and its site carries MuleSoft branding and promotes MuleSoft's Anypoint Platform for broader integration. It runs as a standalone web application or, on paid plans, as a Salesforce Canvas app installed in the org as a managed package, where administrators control which users can run it and which objects they can reach.

The product treats each saved operation as a task. Tasks can be kept and reused without limit, and scheduled tasks run hourly, daily, weekly or monthly with start and end dates. Imports and exports can read from or write to Dropbox, Box and FTP, so a recurring file drop can feed Salesforce without a manual upload. The pricing page also lists related objects and attachments, with limits of 1,000, 10,000 and 100,000 per plan, and file size limits for attachments of 10 MB, 50 MB and 100 MB.

Pricing is per dataloader.io user per month. The Free plan covers 10,000 records a month, one scheduled task and one Salesforce connection, and free tasks expire after 30 days. Professional costs $99 per user a month with 100,000 records, 50 scheduled tasks and email support. Enterprise costs $299 per user a month with unlimited records, unlimited scheduled tasks, SFTP and up to 10 Salesforce connections for moving data between orgs.

Data Loader is Salesforce's own client application for bulk import and export. It inserts, updates, deletes and exports Salesforce records. On import it reads from CSV files or a database connection, and on export it writes CSV files. Salesforce documents it as available in the Enterprise, Performance, Unlimited and Developer editions, and the download page offers versions for macOS and Windows, with Salesforce stating that only the latest version is supported.

There are two ways to run it. The wizard interface handles interactive jobs with drag-and-drop field mapping and a built-in CSV viewer. The command-line interface, available on Windows only, reads configuration, data sources, mappings and actions from files so a job can run unattended on a schedule. Salesforce lists support for all objects, including custom objects, and detailed success and error log files in CSV format.

Volume is the main reason administrators use it. The download page states support for files of up to 5 million records with Bulk API and 150 million records with Bulk API 2.0. If the installed version is newer than the org's API version, Data Loader falls back to the previous API version automatically. Exports are built from a SOQL query, and Salesforce notes that attachments cannot be exported this way; it points to the weekly export feature instead. Data Loader moves files in and out of Salesforce and does not replicate into a warehouse.

MuleSoft Anypoint Platform is an integration and API management platform owned by Salesforce. Its Anypoint Connector for Salesforce lets a Mule application call the Salesforce SOAP, REST, Bulk and Streaming APIs depending on the operation, and react to Salesforce events such as objects, topics, documents or channels being added, changed or deleted. Apex classes can also be called from the connector.

MuleSoft's documentation lists the common use cases as ERP integration, for example syncing Salesforce with SAP or NetSuite through the matching connectors; aggregation of data from several Salesforce orgs, other CRMs and databases into one instance; legacy modernisation, such as migrating data into Salesforce after an acquisition; and connecting Salesforce with ServiceNow, Marketo and similar applications. The connector does not expose every Salesforce API operation and does not cover the Chatter or Tooling APIs.

Building with it assumes developer skills. Users need to know the Salesforce APIs, Mule runtime and either Anypoint Studio or Anypoint Code Builder, and need a Salesforce consumer key and secret. The pricing page lists Integration Starter, covering API and integration design, deployment and low-code tools; Integration Advanced, which adds advanced monitoring, multi-cloud clustering and hybrid deployment; and an API Management Solution. All three are priced on request. The pricing page also lists Agentforce and Einstein connectors for connecting Salesforce AI features to other systems.

Skyvia is a no-code cloud data platform with a dedicated Salesforce connector page. It lists Salesforce integration scenarios in every direction: one-way or bidirectional sync with other sources, ETL or ELT of Salesforce data into a data warehouse, and reverse ETL that loads enriched warehouse data back into Salesforce. Supported Salesforce data includes contacts, leads, opportunities, attachments, accounts and custom objects.

Four Skyvia tools apply to Salesforce work. Import handles bulk loading of historical data and incremental sync of new and updated records, with lookup and expression mapping. Synchronization runs bidirectional sync between Salesforce and other cloud apps or databases. Data Flow joins several sources through lookups, applies multistage transformations and writes to multiple destinations in one run. Control Flow orchestrates integrations in sequence or in parallel with error handling. The connector list on the Salesforce page runs to well over a hundred applications and databases, including Dynamics 365, HubSpot, Google BigQuery, Amazon Redshift and Databricks.

Data Integration plans are priced by records per month and scheduling frequency. Free covers 10,000 records a month and two scheduled integrations that expire after 30 days. Basic costs $99 a month or $79 billed annually. Standard costs $199 or $159 annually with hourly runs and 50 integrations. Professional costs $499 or $399 annually with per-minute runs and unlimited integrations. Enterprise adds a dedicated virtual machine and unlimited records on a custom annual package. Skyvia also sells a separate Salesforce backup product.

Hevo Data is a managed pipeline product, and Salesforce is one of its documented SaaS sources. Since 3 February 2025, Hevo's Salesforce integration has used Salesforce Bulk API 2.0 by default. New pipelines can no longer be created on the older Bulk API integration, although existing ones continue to run and can be upgraded from the pipeline overview page.

Setup connects either a production or a sandbox environment through OAuth. Hevo states that accounts with single sign-on enabled are not supported, and its documentation recommends creating a dedicated user with API Enabled access, access to the objects being ingested and an SSO bypass. The historical sync duration defaults to three months and can be extended to all available data, but it cannot be changed after the pipeline is created. New objects created in Salesforce later can be included automatically or held in a skipped state.

After the historical load, Hevo fetches only records modified since the last sync by checking SystemModstamp, LastModifiedDate, CreatedDate and LoginTime in that order. Objects without any of these fields are fully reloaded on each run. Hevo recommends a destination table prefix so Bulk API 2.0 tables do not collide with tables from an older Salesforce pipeline.

Pricing is based on events, meaning records inserted, updated or deleted in the destination. Starter costs $299 a month, or $239 annually, at 5 million events. Professional costs $849 a month, or $679 annually, at 20 million events.

CData Sync is a data replication product that runs as CData's cloud service, in a self-hosted cloud or on-premises. It moves data from applications, databases and APIs into warehouses, lakes and databases, and can push data back into SaaS applications for reverse ETL. CData states that pricing is based on connections, with no fees for row counts, storage or usage spikes.

Salesforce is documented as both a source and a destination. The source documentation describes a Salesforce-specific issue: when the definition of a formula field changes, Salesforce does not update the record's SystemModStamp, so incremental replication that relies on that timestamp misses the change. Sync addresses this with two schemas. The Salesforce schema replicates standard fields incrementally, with a setting to include or skip formula fields. The Salesforce_Formulas schema contains only the record ID and formula fields, truncates and reloads them on each run, and can add new formula fields automatically.

As a destination, Salesforce supports several authentication methods, including OAuth, OAuth JWT, OAuth PKCE, OneLogin, PingFederate, Okta, ADFS and Azure Active Directory, plus sandbox connections and a configurable API version.

The Standard licence costs $7,999 a year for up to five standard connections and 100 million processed rows a month. Unlimited rows, premium connectors, change data capture, ETL or ELT transformation, role-based access control, single sign-on and REST API access are quoted. Marketplace images on AWS and Azure start at $2.74 an hour.

Jitterbit is a low-code integration platform as a service (iPaaS) with a dedicated Salesforce integration offering. Its Salesforce page groups use cases by business process rather than by data movement. Lead-to-revenue integration syncs Salesforce CRM with ERP, CPQ, billing and payment systems to speed up quote-to-cash and order processing. Connected commerce links Salesforce Commerce Cloud and Shopify with ERP and CRM records. Marketing integration connects Salesforce Marketing Cloud and Marketo with the CRM, and support integration covers Service Cloud, Jira and Zendesk.

The platform builds integrations in Studio, a visual designer with drag-and-drop prebuilt connectors, and adds a Management Console, a Connector Builder and a Connector SDK. Integrations run on Jitterbit cloud agents or on private agents installed in the customer's environment, which matters when Salesforce must talk to an on-premises ERP or database. API Manager is an add-on for publishing custom APIs.

Pricing is quote-based and organised by connections. Standard includes two to three connections, Professional five and Enterprise eight or more, with additional connections available as add-ons. Every product requires an annual contract. Support response targets are 48 hours on Standard, 24 hours on Professional and 6 hours on Enterprise. Jitterbit also offers enterprise licence agreements and discounts for nonprofit and education buyers.

Celigo is an integration platform whose Salesforce page focuses on business processes. Listed use cases include quote-to-cash, converting Salesforce quotes and opportunities into sales orders or invoices in an ERP; ecommerce integration that syncs customer and order data; sales and support integration; subscription and contract syncing for renewals and billing; marketing automation; a customer 360 view across Salesforce and ERP; project management; and financial forecasting. The page also lists syncing Salesforce data into warehouses such as Snowflake for reporting.

Prebuilt Salesforce integrations are a central part of the offer. Celigo names Shopify, Amazon, HubSpot, Snowflake and NetSuite as ready-made pairings, alongside a general builder for custom flows between Salesforce and other applications. All editions include Celigo's AI capabilities, including an Agent Builder and an MCP server.

The pricing page describes three editions. Standard targets the most critical integrations, with MFA and role-based access and optional SSO. Professional adds API Builder, Lookup Cache, SSO with auditability and a sandbox. Enterprise includes unlimited endpoints, maximum performance and the fastest support SLAs. Celigo charges by endpoints and flows rather than tasks, data volume or vCores, and its FAQ states that it does not offer volume-based pricing and charges no overage fees. Buyers add endpoints or flows through an account manager. No edition has a published price.

Hightouch is a reverse ETL and composable customer data platform. Instead of extracting Salesforce data, it reads models from a warehouse such as Snowflake or Google BigQuery, or from databases, files and BI tools, and writes the results into business applications. Salesforce is one of its documented destinations, covering Sales Cloud, Service Cloud, Financial Services Cloud and Health Cloud.

The Salesforce destination supports three sync types. Standard and custom objects such as accounts and contacts can be synced in insert, update, upsert or archive mode. Picklist field metadata can be synced in all mode, which overwrites the picklist values each run. Platform events can be published in insert mode, so Salesforce flows, processes and Apex triggers can start from warehouse data.

The documentation sets out the Salesforce checks needed before a sync. The matching field must be an external ID and unique to appear as a record-matching option. Target fields must be editable rather than formula or read-only, and the authenticating user needs edit access or the sync will error. Model column types may need casting to match Salesforce field types.

Basic Reverse ETL is a free tier with up to two active syncs, unlimited destinations and unlimited seats. Paid products are usage-based with no published figures. The pricing page lists SOC 2, HIPAA, GDPR and ISO 27001 compliance and SAML SSO.