Fivetran Connectors

Managed ELT connectors that move data from SaaS apps and DBs into warehouses.

Reviewed by 7wData

On this page

Publisher review

Fivetran is a managed ELT platform that automates data movement from 700+ sources—SaaS applications, databases, files, APIs—into data warehouses and lakes. The product handles schema detection, incremental syncing, and basic transformations automatically, requiring no engineering effort to deploy or operate. Founded in 2012, Fivetran dominates the managed connector space by covering breadth (SaaS tools, cloud databases, legacy systems) that would take months to build in-house.

The platform is strongest for organizations that need to ingest data from dozens of disconnected sources without maintaining custom connectors. Fivetran's dbt integration lets teams automate transformations immediately after data arrives, closing the ELT pipeline within the platform. However, the product's real-time story is complicated.

Standard tier guarantees 15-minute sync intervals; achieving 1-minute latency requires Enterprise pricing at 2–3x cost. For change data capture at scale, Fivetran offers HVR, a self-hosted agent product that adds operational overhead (provisioning, scaling, maintenance) and covers only 30 source types. In 2026, the main tension is cost.

Fivetran's usage-based model charges per Monthly Active Row (MAR), and two structural changes have made it expensive: a March 2025 move to per-connector billing eliminated bulk discounts, and a January 2026 $5 minimum per connection catches low-volume pipelines. More problematically, Fivetran's automatic schema normalization—splitting nested JSON into multiple rows to flatten semi-structured data—inflates MAR counts unexpectedly. A single Shopify order, for example, becomes four rows; customers pay for all four.

The platform also applies schema changes automatically without user control, making future migrations uncertain. Community feedback surfaces reliability gaps: while popular connectors work seamlessly, less common integrations encounter issues, and organizations report slow support. The trade-off is real: Fivetran trades operational simplicity for cost visibility and customization control.

Get the AI & data signal, daily.

335k+ subscribers read this every morning. One email, both newsletters. Unsubscribe anytime.

How it works

  1. 700+ pre-built connectors

    Covers SaaS platforms (Salesforce, HubSpot, Stripe), databases (PostgreSQL, MySQL, SQL Server, MongoDB), ERPs (SAP, Oracle), and data APIs without custom code.

  2. Automated schema detection and normalization

    Automatically maps source schemas to warehouse-compatible tables and normalizes nested data into flat structures; no DDL required.

  3. dbt Core and dbt Cloud integration

    Triggers dbt jobs automatically after data syncs complete; Fivetran hosts dbt Core execution or orchestrates dbt Cloud runs for downstream transformations.

  4. Batch and log-based CDC

    Standard tier offers 15-minute sync intervals via batch polling; Enterprise tier supports 1-minute latency via transaction log reading (WAL/binlog).

  5. HVR self-hosted agent for high-volume databases

    Separate agent product installed on source database servers for real-time CDC; covers ~30 database types and requires infrastructure provisioning.

  6. Managed Data Lake and reverse ETL

    Builds Iceberg and Delta Lake tables automatically; Activations feature pushes data back to business tools (Salesforce, Marketo, etc.).

  7. Hybrid and private deployment

    Cloud-native SaaS option and self-hosted deployment models for organizations requiring data residency or air-gapped environments.

Strengths and trade-offs

Strengths

  • Breadth of pre-built connectors eliminates months of custom integration work; low operational overhead for teams uncomfortable owning data pipelines.
  • Automated schema handling reduces schema management burden; integrates tightly with dbt ecosystem for codified transformations.
  • Enterprise compliance certifications (SOC 2, HIPAA, GDPR, PCI DSS) and audit trails support regulated industries without additional tooling.

Trade-offs

  • Cost is the dominant complaint; per-connector billing (March 2025) and $5 minimum per connection (January 2026) have driven 4–8x increases; hidden normalization costs inflate row counts unexpectedly (single nested records split into multiple rows).
  • Latency is batch-first; 15-minute sync intervals are standard; 1-minute latency requires Enterprise tier at 2–3x cost; HVR (real-time CDC) is a separate product requiring self-hosted infrastructure.
  • Limited customization; schema changes are applied automatically without user control; reliability gaps with niche connectors; support response times criticized in community feedback.

Pricing context

Fivetran offers a free tier (500K Monthly Active Rows monthly) for evaluation, then usage-based pricing on Standard, Enterprise, and Business Critical tiers. Billing is per-connector based on data volume; a $5 minimum monthly charge applies to each connection. Standard tier MAR pricing follows a declining schedule: $2.50 per million rows (0–5M), $2.00 (5–20M), $1.50 (20–100M), $1.00 (100M+).

Real-world costs range from ~$500/month for light usage (4 connectors, median volumes) to $20,000+ for enterprise deployments. Transformation runs beyond the free 5,000/month tier cost $0.01 each. January 2026 brought the $5 base charge; March 2025 switched to per-connection billing (eliminating prior account-level discounts).

Annual contracts offer up to 22% savings, starting at 5% discount. The consumption model creates budget uncertainty for growing data volumes; organizations typically require custom licensing agreements for predictable costs.

Getting started with Fivetran Connectors

  1. Sign up for Fivetran trial

    Create a Fivetran account or start a free trial with 500,000 monthly active rows included. Once verified, you'll land in the dashboard where you can select your first data source and destination warehouse.

  2. Choose and configure your first connector

    Browse the 700+ pre-built connectors and select the one matching your data source: Salesforce, Stripe, PostgreSQL, or another platform. Enter the required credentials for that source. Fivetran will test the connection.

  3. Point connector to your data warehouse

    Select your destination warehouse: Snowflake, BigQuery, Redshift, or another supported platform. Provide the warehouse credentials. Fivetran automatically detects the source schema and creates matching tables in your warehouse without manual DDL work.

  4. Trigger and monitor your first sync

    Click Start Sync and monitor the initial data movement in the Fivetran dashboard. Watch the row counts and sync duration. Once complete, verify that data arrived correctly in your warehouse tables.

  5. Schedule regular syncs and set alerts

    Set the sync frequency: Fivetran defaults to 15-minute intervals on Standard tier. Configure dbt Core or dbt Cloud if you plan to transform data post-sync. Enable cost and failure alerts in settings.

Frequently Asked Questions

What is Fivetran?

Fivetran is a managed ELT platform that automates data movement from 700+ sources into data warehouses and lakes. It handles schema detection, incremental syncing, and basic transformations automatically, requiring no engineering effort to deploy or operate. Founded in 2012, it dominates managed connectors.

How many data sources can Fivetran connect to?

Fivetran covers 700+ pre-built connectors spanning SaaS platforms (Salesforce, HubSpot, Stripe), databases (PostgreSQL, MySQL, MongoDB), ERPs (SAP, Oracle), and data APIs. This breadth eliminates months of custom integration work and appeals to organizations needing to ingest data from multiple disconnected sources without maintaining custom connectors.

What does Fivetran cost?

Fivetran charges based on Monthly Active Rows (MAR) with a free tier offering 500K MAR monthly. Standard tier pricing ranges from $2.50 per million rows (0–5M) down to $1.00 (100M+), plus a $5 monthly minimum per connection. Real-world costs range from ~$500/month for light usage to $20,000+ for enterprise deployments.

Why is Fivetran becoming more expensive?

Fivetran's costs have risen due to structural billing changes: March 2025 shifted to per-connector billing (eliminating prior bulk discounts), and January 2026 introduced a $5 minimum per connection. Additionally, automatic schema normalization inflates row counts unexpectedly—a single nested record splits into multiple rows, increasing Monthly Active Rows charges unpredictably.

How fast does Fivetran sync data?

Standard tier guarantees 15-minute sync intervals via batch polling. Achieving 1-minute latency requires Enterprise pricing at 2–3x cost. For real-time change data capture at scale, Fivetran offers HVR, a separate self-hosted agent covering ~30 database types, but adds operational overhead for provisioning and maintenance.

What are Fivetran's main limitations?

Fivetran applies schema changes automatically without user control, limiting customization. Less common connectors encounter reliability issues, and community feedback criticizes slow support response times. The platform also trades operational simplicity for cost visibility and control, making budget forecasting difficult for growing data volumes.

Alternatives in this category

Integrations

Snowflake BigQuery Databricks Redshift

How Fivetran Connectors compares

Direct head-to-head against 3 competitors. Picked by 7wData.

This tool

Fivetran Connectors

Pricing
Fivetran offers a free tier (500K Monthly Active Rows monthly) for evaluation, then usage-based pricing on Standard, Enterprise, and Business Critical tiers. Billing is per-connector based on data volume; a $5 minimum monthly charge applies to each connection. Standard tier MAR pricing follows a declining schedule: $2.50 per million rows (0–5M), $2.00 (5–20M), $1.50 (20–100M), $1.00 (100M+). Real-world costs range from ~$500/month for light usage (4 connectors, median volumes) to $20,000+ for enterprise deployments. Transformation runs beyond the free 5,000/month tier cost $0.01 each. January 2026 brought the $5 base charge; March 2025 switched to per-connection billing (eliminating prior account-level discounts). Annual contracts offer up to 22% savings, starting at 5% discount. The consumption model creates budget uncertainty for growing data volumes; organizations typically require custom licensing agreements for predictable costs.
Target
Fivetran is a managed ELT platform that automates data movement from 700+ sources—SaaS applications, databases, files, APIs—into data warehouses and lakes.
Deployment
cloud
Strength
Breadth of pre-built connectors eliminates months of custom integration work; low operational overhead for teams uncomfortable owning data pipelines.
Watch for
Cost is the dominant complaint; per-connector billing (March 2025) and $5 minimum per connection (January 2026) have driven 4–8x increases; hidden normalization costs inflate row counts unexpectedly (single nested records split into multiple rows).

Airbyte Cloud

Pricing
Standard $10/month base plus $2.50/credit overages; Plus $25,000/year minimum; Enterprise custom.
Target
Data engineering teams wanting open-source connector breadth with a managed cloud deployment option.
Deployment
SaaS or self-hosted open-source
Strength
600+ open-source connectors; community contributions and AI-assisted builder outpace proprietary connector catalogues.
Watch for
Community-contributed connectors regularly fail or stall in production; beta connectors lack proper state recovery on sync crashes.

Stitch Data

Pricing
Standard $100/month (5-300M rows); Advanced $1,500/month; Premium $3,000/month; all billed annually.
Target
SMB data teams needing simple, fast ELT setup with flat-rate monthly pricing.
Deployment
SaaS
Strength
140+ connectors with fastest time to first sync; flat-rate tier pricing versus per-row billing surprises.
Watch for
Acquired by Talend, then Qlik (2023); new users directed to Qlik Talend Cloud; feature development has stalled since acquisition.

Hevo Data

Pricing
Free tier 1M events/month; Starter from $239/month; Professional from $679/month; Business Critical custom.
Target
Mid-market data teams seeking transparent event-based pricing to avoid Fivetran nested-record row inflation.
Deployment
SaaS
Strength
24/7 expert support at every paid tier; auto-healing pipelines restore failed syncs without manual intervention.
Watch for
~150 connectors versus Fivetran's 700+; enterprise compliance certifications less mature; smaller community for troubleshooting edge cases.

User reviews

No user reviews yet. Be the first to write one.

Sources

Reporting on this tool draws on these publicly available sources.

  1. www.fivetran.com — Free tier allowances (500K MAR), Standard/Enterprise/Business Critical tiers, $5 base charge per connection, dbt Core integration, 700+ connectors, 15-minute sync intervals, 200+ activation destinations
  2. www.integrate.io — MAR billing mechanics, per-connection March 2025 changes, transformation fees ($0.01 per run after 5,000 free), real-world cost ranges ($48k–72k annually for 10M rows), January 2026 $5 minimum charge, data deletion billing
  3. estuary.dev — Batch-only CDC limitations, 15-minute vs 1-minute latency tiers, database stress from CDC, schema change control limitations, row normalization cost issues, reliability gaps with niche connectors
  4. www.g2.com — 35% of recent reviewers cite pricing as primary concern, 4–8x cost increases reported, data discrepancies on reliability, auto-applied schema changes, limited customization feedback (community consensus)
  5. cdcalculators.com — MAR tiered pricing structure ($2.50/M down to $1/M at scale), per-connector billing mechanics, typical cost scenarios
  6. hevodata.com — January 2026 $5 base charge on standard connections, March 2025 per-connector billing transition, hidden normalization costs, semi-structured data row inflation examples
  7. fivetran.com — Usage-based pricing (Monthly Active Rows), connection-level cost calculation, spend rate tiers, automatic volume discounts
  8. www.fivetran.com — dbt Core integration, dbt Cloud orchestration, Quickstart data models, transformation automation after syncs