API integration for a UK SME typically costs between 5,000 and 50,000, depending on the number of systems connected, the complexity of data mapping, and whether the sync is real-time or batch. A single integration between two systems with well-documented APIs starts at around 5,000. A full integration project connecting five or more systems with real-time bidirectional sync typically costs 25,000 to 50,000.
We build custom API integrations and middleware for UK businesses that need their systems to work together without manual data entry. This guide breaks down the costs so you can budget accurately.
What Are the Cost Ranges for Different Integration Types?
Integration projects vary significantly in complexity. Here are the typical cost ranges:
- Single one-way integration: 5,000 to 10,000. Data flows from one system to another on a schedule. Example: syncing new CRM contacts to an email marketing platform.
- Single bidirectional integration: 10,000 to 20,000. Data syncs both ways, with conflict resolution. Example: syncing contacts between CRM and accounting, with updates in either system reflected in both.
- Real-time integration: 15,000 to 30,000. Data syncs instantly via webhooks or streaming, not on a schedule. Example: order data flowing from e-commerce to fulfilment in real-time.
- Multi-system middleware: 25,000 to 50,000. A central integration layer connecting three to five systems, with data transformation, routing, and error handling.
- Full integration platform: 40,000 to 80,000. A custom integration hub connecting all business systems, with monitoring, alerting, and a management dashboard.
What Drives the Cost of API Integration?
Six factors determine the cost of an integration project:
- Number of systems: Each additional system adds design, development, testing, and maintenance work. The complexity grows non-linearly because each system can interact with every other.
- API quality: Well-documented APIs with stable endpoints, good authentication, and clear data models are faster and cheaper to integrate. Poorly documented or frequently changing APIs add significant time.
- Data mapping complexity: If the data models between systems are similar, mapping is straightforward. If they are different (different field names, data types, relationships), custom transformation logic is needed.
- Sync direction and timing: One-way batch sync is the simplest and cheapest. Bidirectional sync requires conflict resolution. Real-time sync requires webhooks or streaming infrastructure.
- Error handling: What happens when an API is down, returns an error, or sends unexpected data? Solid error handling with retries, alerts, and manual intervention options adds cost but is essential for reliability.
- Data volume: High-volume integrations need performance optimisation, batching, and sometimes different technical approaches than low-volume sync.
Should You Build Custom Integrations or Use Middleware?
Middleware platforms like Zapier, Make (formerly Integromat), and Workato offer no-code integration between popular SaaS tools. For simple workflows, they are faster and cheaper than building custom.
Use middleware when:
- Connecting standard SaaS tools with pre-built connectors
- Simple trigger-action workflows (when X happens in system A, do Y in system B)
- Low data volume (under a few thousand records per month)
- No complex data transformation needed
Build custom when:
- Connecting custom or legacy systems without pre-built connectors
- Complex data transformation or business logic is required
- Real-time sync is needed
- Data volume is high enough that middleware per-task pricing becomes expensive
- You need full control over error handling, retry logic, and monitoring
Middleware costs scale with usage. A business running 50,000 tasks per month on Zapier can pay 500 to 1,500 per month. Custom integration has a fixed upfront cost and minimal ongoing cost, making it cheaper at scale.
What Is the Process for Building API Integrations?
A typical integration project follows these steps:
- API discovery (1 to 3 days): Review API documentation for each system. Identify available endpoints, authentication methods, rate limits, and data models.
- Data mapping design (2 to 5 days): Map fields between systems. Identify data transformations, default values, and conflict resolution rules.
- Build and test (1 to 4 weeks): Develop the integration, test with real data, and handle edge cases and errors.
- Monitoring setup (1 to 2 days): Set up alerts for failed syncs, monitoring dashboards, and logging for troubleshooting.
- Deployment and validation (1 to 2 days): Deploy to production, run initial full sync, and validate data accuracy across systems.
The total timeline for a single integration is one to four weeks. A multi-system project takes six to twelve weeks.
What Are the Ongoing Costs?
API integrations need maintenance. External APIs change endpoints, authentication methods, and data models over time. Budget for:
- Monitoring and maintenance: 10 to 15 percent of build cost annually. For a 15,000 integration, that is 1,500 to 2,250 per year.
- API changes: When a vendor updates their API, the integration may need updates. Budget one to two days of developer time per quarter for each external API.
- Middleware subscriptions: If using middleware, the subscription cost continues. For custom integrations, hosting costs are minimal (50 to 200 per month).
For businesses with multiple integrations, a central monitoring dashboard and alerting system is worth the investment. It catches failures before they cause data inconsistencies.
Need help connecting your systems? Book a free discovery call to discuss your integration needs, or see our services for the full range of what we offer. For more on system integration concepts, see our guide on what is system integration.