SPSync vs Fivetran

Choose SPSync when SharePoint lists are a big source and a Monthly Active Rows bill is hard to forecast; choose Fivetran when lists are one connector among many in a managed pipeline you already pay for.

Documentation reviewed 23 October 2026

Where Fivetran fits

• Fully managed, minimal setup
• reliable, enterprise-grade schema handling
• wide destination list including Snowflake, BigQuery, Databricks and more
• usage-based pricing that follows distinct rows changed each month
• lookup columns arrive as an id column you can join on
• permissions sync for replicated SharePoint files on Advanced setup

Where it breaks

• Pricing is usage-based on Monthly Active Rows, so cost follows how many distinct rows change each month
• multi-choice, people and managed metadata columns are not documented as split into child tables
• runs in Fivetran's cloud, or in your environment with Hybrid Deployment on Enterprise and Business Critical plans, rather than as a desktop app or CLI-first workflow
• documented for SharePoint sites, with no on-prem SharePoint support documented
• its agentic AI material for SharePoint is about documents and files, with list rows outside that story

Feature by feature

With Fivetran, the pipeline is somebody else’s problem. You point it at a source, pick a destination, and a managed connector handles scheduling, retries and schema drift with no infrastructure to maintain. A team already standardised on Fivetran across dozens of SaaS sources can add SharePoint as one more connector and keep Fivetran’s reliability and support behind it.

Monthly Active Rows and the list connector

Fivetran prices on Monthly Active Rows. Its documentation says a row is counted once a month however many updates it receives, so cost follows how many distinct rows change in a month, whatever the number of edits. Fivetran documents its Microsoft Lists connector as syncing each list to a table, and its separate SharePoint connector as syncing files. A lookup column arrives as column_name_lookup_id, holding the id of the item it points at, so lookups can be joined. We found no documentation describing multi-choice, people or managed-metadata columns split into child tables (checked 2026-10-23), so expect to model those in SQL or dbt.

Tables shaped by SharePoint’s column types

SPSync replicates each list into tables built around SharePoint’s column types. Lookups become key columns with the display value stored beside them, multi-lookup and multi-choice columns become child tables, and people resolve into a shared sp_users table. Delta sync runs every few minutes by default with a reconcile (weekly by default), and licensing is per installation and per target family, with no count of active rows.

Teams most often move when SharePoint is a significant source in its own right and a usage-based line on the Fivetran bill is hard to budget for.

Is Fivetran’s SharePoint data ready for AI agents?

“We want an agent to answer from our SharePoint data, and the vendor says it is agentic AI ready.” Fivetran’s own guide to SharePoint and agentic AI is about documents. It moves documents, spreadsheets, images and text files, with their metadata, into a warehouse or data lake, and leaves dbt to model them into tables an agent can use. On Advanced setup it can also sync permissions for those replicated files, so an agent can respect who may see which document. For a team whose agents answer from policies and contract files, that is a reasonable foundation, and Fivetran handles the plumbing.

List rows sit outside that story. They travel through the Microsoft Lists connector described above, on the same Monthly Active Rows billing, and the guide does not cover them.

SPSync is built for the list side. It keeps one SQL copy of your lists and library metadata, refreshed every few minutes, and never writes back to SharePoint. An agent queries that copy through a database MCP server you choose, or through spsync mcp serve, SPSync’s own read-only server, signing in with a read-only database login, so it never holds a Microsoft Graph permission. If you turn embeddings on, SPSync writes vectors on each sync into pgvector or Azure AI Search, and nowhere else today. SPSync ships no agent of its own and copies metadata, not file contents. SharePoint item permissions stay in SharePoint, so row-level security in the database decides what each agent sees.

Switching the lists from Fivetran

1. Note the site and lists your Fivetran Microsoft Lists connector reads, and the destination schema it writes to.
2. Sign in to SharePoint through SPSync and choose those lists.
3. Pick a destination database, run the first full load, then set the delta schedule.
4. Compare the SPSync tables with Fivetran’s output, especially lookup and multi-value columns, and update downstream queries to the new schema.
5. Pause or delete the connector in Fivetran when you are satisfied, to stop usage charges for it.

SPSync as a Fivetran alternative for SharePoint lists

Moving the lists to SPSync takes one connector out of the Monthly Active Rows count and leaves the rest of the Fivetran account alone. The new tables carry lookups as keys and multi-value columns as child tables, and the licence stays the same however many rows change.

Other alternatives

• Airbyte: an open-source platform; its SharePoint Lists Enterprise connector is no longer sold, and the Marketplace Microsoft Lists connector runs full refreshes.
• Coupler.io: scheduled Microsoft Lists exports into BigQuery, Snowflake, PostgreSQL and BI tools on flat-priced plans.
• Peliqan: a hosted warehouse and BI layer that takes SharePoint lists alongside its other sources.
• Estuary: a streaming pipeline whose SharePoint connector captures document library files, with list items left out.