Make integration cost

What drives integration cost

Modules or HTTP

Custom API calls take longer than built-in modules.

Authentication

OAuth and token refresh add setup.

Direction

Two-way sync needs conflict rules.

Volume and paging

Large data needs paging and batching.

Safeguards

Verification and duplicate checks add design.

Ask for operations per typical run and per peak run.

Integrations compared

How integrations are priced

Operations and running cost

Each module run per bundle is an operation, so an integration moving a thousand records through four modules uses about four thousand operations per run, plus any trigger checks. Filtering early and avoiding unnecessary lookups reduce this. Check Make's current plans against your estimates.

Two-way sync

When both systems update each other, the integration must prevent loops and decide which change wins. That doubles the flows and adds logic. One source of truth per field is usually cheaper.

Example: reducing operations in an integration

Comparing quotes

Check which APIs were reviewed, direction, volume handling, verification, duplicate prevention, field map, failure tests, operation estimate, documentation and support.

Keeping the cost down

Choose one source of truth per field, sync only changed records, prefer apps with modules, move only needed fields and provide API access promptly. Provide realistic sample data early.

Custom API work

Poorly documented APIs, unusual authentication or strict limits multiply effort. A feasibility review reduces risk for both sides. Ask which assumptions the quote makes about each API.

Support

APIs change and credentials expire. Budget for fixes through a support arrangement or hourly rate. Keep a list of which scenarios use which apps, so changes are traced quickly.

Judging the value

Compare re-keying removed, errors avoided and faster data with build, operation and support costs over a year. Include the time spent fixing errors by hand today.

Authentication

API keys are quick. OAuth connections may need app registration, approval from the app provider and token refresh handling. Some systems take days to approve API access, which can delay projects.

Volume and paging

Moving thousands of records means paging through results, batching writes, respecting rate limits and resuming safely after failures. Share volumes, including peaks and bulk imports, so quotes and operation estimates are realistic.

Testing

Testing duplicates, large batches, missing fields and outages takes time, especially without sandbox accounts, but prevents double invoices and lost records in live systems.

Documentation

A field map and short notes take little time and make later changes far cheaper, especially if someone else maintains the integration.

Reconciliation

A scheduled check that record counts or totals match between systems adds a little to the build and catches silent drift. It is worth including for finance and stock data.

Custom webhooks

Receiving data through custom webhooks avoids polling operations and delivers data immediately, but needs verification, duplicate handling and fast responses. These add build time, which is usually recovered through lower operation use.

Field mapping and lookups

Formatting fields for the destination and looking up existing records before creating new ones add a few modules per flow. They prevent messy data and duplicates, which cost far more to clean up later.