At a glance
- There is no per-row or per-connector charge: Ingest is priced on the compute that runs your syncs.
- Each connection runs next to Fivetran until schemas, keys and totals match.
- Rig engineers do the build and reconciliation, and your team signs off each wave.
- A Fivetran connector can stay where Rig does not cover the source.
Records, logic and owners
Source to target definitions
Counts, totals and access
Sign-off and rollback plan
What Fivetran charges for today
Fivetran prices each connection on monthly active rows (MAR), and each connection follows its own cost curve, with a spend rate that falls as usage rises. Fivetran adds a $5 base charge to standard connections with 1 to 1M MAR, bills transformations per model run above the first 5,000 runs a month, and says an annual contract saves up to 22%. Model your connections in the Fivetran cost calculator before you compare quotes. Rig Ingest is priced on the compute that runs your syncs, with no per-row or per-connector charge, which typically works out 10 times cheaper or more. Rig prices the platform and delivery after scoping your sources.
Map the data and the logic it carries
The table is a starting inventory, and the target edition, customisations and API access decide the final mapping. Check each object even when the name matches, because an object with the same name can behave differently in the target.
| Source | Target or retained output | Checks and known traps |
|---|---|---|
| Connections and credentials | Equivalent target connections | Confirm source, destination, auth and API coverage. |
| Tables, keys and sync state | Target schema and incremental loads | Compare deleted rows, history and schema drift behaviour. |
| dbt models and downstream reports | Retained warehouse models | Check schema names, data types and refresh dependencies. |
How a migration runs in Rig
Rig maps the systems and data across your estate, connects the source and target, and maps one schema to the other with previews of real records. Rig engineers agree the objects and history in scope with your team, prepare the destination and rehearse the load before they reconcile the results. Sign-off, cutover and follow-up tasks sit in the same migration workspace.
Agree a timeline against the actual scope
The table gives a planning allowance for each phase, and the assessment turns it into a dated plan for the whole project. Large estates need several waves, and application configuration can add time.
| Phase | Initial planning allowance | Exit check |
|---|---|---|
| Assessment | 3 to 5 working days | Owners, access and inventory agreed |
| Sample and mapping | 5 to 10 working days | Sample records and exceptions reviewed |
| Build and parallel validation | 2 to 6 weeks per wave | Counts, totals and permissions accepted |
| Final sync and cutover | 1 to 3 working days per wave | Business owner signs off; rollback remains available |
Keep the evidence for each migration wave
For each wave you receive the source-to-target mappings, a reconciliation report, an exception log and a cutover checklist. The checks cover duplicates, missing relationships, deleted records, time zones and permissions as well as headline totals, and the old system stays available until the agreed retention and rollback checks pass.
Compare the alternatives before migrating
A cheaper pipeline only helps if it covers your sources and refreshes as often as the business needs. The options below differ in coverage, operating work and pricing, so run the same connection through each one before you choose.
| Option | Pricing basis | Work to plan | Where it fits |
|---|---|---|---|
| Fivetran | Monthly active rows, with plan and connection terms. | Own the warehouse models and downstream tests. | Teams needing broad managed source coverage. |
| Airbyte | Self-hosted infrastructure; cloud credits or capacity by edition. | Self-hosting includes upgrades, security and connector operations. | Teams that want to self-host and develop connectors. |
| Estuary | Data moved and active connector instances; enterprise terms vary. | Choose capture and delivery modes, and monitor volume and connector costs. | Operational workloads that need fresh changes, replay or multiple destinations. |
| Stitch | Row-based plans and plan-specific add-ons. | Maintain transformations and downstream reporting. | Teams with straightforward supported loading requirements. |
| Hevo | Event quotas with plan-specific terms. | Manage destination models and monitor business totals. | Teams evaluating a managed pipeline service. |
| Matillion | Credits with product-specific consumption rules. | Build and operate transformations and orchestration. | Teams with substantial transformation requirements. |
| Rig Ingest with delivery | Priced on compute (the servers that run the syncs), with no per-row or per-connector charge; typically 10 times cheaper or more than row-based pricing. Platform and delivery are priced separately. | Agree sources, tables, refresh schedules and reconciliation checks. | Routine warehouse ingestion where scheduled freshness meets the business need. |
Where Rig may fall short
A migration that only needs a supported CSV import may be simpler to do yourself. Sub-minute operational replication, a full ERP implementation and specialist application redesign need a separate scope or another provider. Where Rig Ingest moves the data, confirm connector coverage for each source before you commit.
Inside Rig Migrate
Run the migration in Rig, from the systems map to cutover
Rig natively maps systems and data across your IT estate, connects source and target systems, scopes the move and maps one schema to another. The same migration workspace tracks destination preparation, rehearsals, reconciliation, sign-off and the work after cutover, so your team can review the decisions and progress in one place.
Map the IT estate
Map systems, data and integration flows across the IT estate. See which reports, processes and teams depend on the systems you plan to change.
Set up the migration
Create the migration in Rig and record the source, destination and work to deliver. Keep the migration tasks and progress together.
Connect and discover
Connect the source and target systems, then inspect their schemas and available data. Confirm access before deciding how each object will move.
Scope the work
Choose the objects, fields and history to migrate. Record what stays, what changes and the dependencies the migration needs to preserve.
Map schema to schema
Map source objects and fields to the target schema. Preview records and review rules to copy, translate, calculate, link, look up and match owners.
Prepare the destination
Track the target preparation needed before loading data, including its configuration and the relationships your mapped records will use.
Rehearse the load
Dry run the whole load and rehearse it in the destination environment. Review the results and resolve issues before the production cutover.
Reconcile and sign off
Compare source and target records, relationships and business totals. Review exceptions with the system owners and record acceptance before proceeding.
Cut over
Work through the agreed cutover tasks in Rig. Coordinate the final load and user switch with the people who own the systems.
Check after cutover
Track the checks after the switch, confirm that reporting and processes work, and resolve outstanding issues before retiring the old setup.

Inspect source values and target previews
The migration workbench shows source fields, target previews and the rule for each mapping. Your team can review translated values, linked records and calculated fields before signing off the mapping and rehearsing the load.

A managed data stack at SURI
SURI had 40 data sources and no data team. One analyst brought them together on Rig, had finance sign off every metric in a 15-minute meeting, and turned a monthly P&L into a weekly one.
Read the SURI case studyCommon questions
How long does it take to replace Fivetran?
The assessment takes 3 to 5 working days and produces a dated plan. Each wave then runs in parallel until your team accepts the reconciliation, so the number of connections, sync modes and downstream models sets the total.
Does Rig charge per row like Fivetran?
No. Rig Ingest is priced on the compute that runs your syncs, with no per-row or per-connector charge. That typically works out 10 times cheaper or more than row-based pricing. Warehouse compute stays on your own bill, or Rig hosting passes it through at cost. Platform and delivery fees are agreed after scoping.
Do our dbt models and reports have to change?
They stay where they are, in your warehouse. During the parallel run Rig checks schema names, data types and refresh dependencies against them, and fixes any difference before cutover.
Can we keep some Fivetran connections?
Yes, where Fivetran covers a source or sync mode that Rig does not. The coverage review records which connections move to Rig and which stay where they are.
How much does a migration cost?
Rig maps the work and agrees a fixed delivery price before the migration starts. Source access, custom logic, history and the number of integrations affect the scope. Platform and third-party licence costs are agreed separately.
How do you check that the data matches?
Rig compares record counts, identifiers, relationships and business totals for each wave. Every exception goes in a log, and the business owner has to accept the results before cutover.
Who does the implementation?
Rig engineers do the work with your system owners, and your team approves the mappings, permissions and acceptance checks. Application configuration and specialist ERP or CPQ work is either written into the scope or assigned to your application partner.
Sources and product details
Vendor information checked on 20 November 2026. Confirm current editions, licence terms and prices with the vendor before buying.