Unified.to
All articles

Replace Polling with Unified Webhooks: Real-Time and Historical Data in One Flow


January 10, 2024

Unified.to_webhooks_enhancements.png

Unified webhooks now support initial backfill and ongoing updates through a single subscription.

This removes the need to:

  • build separate sync jobs for historical data
  • manage polling infrastructure
  • stitch together backfill + incremental pipelines

You can now subscribe once and receive:

  • existing records (initial sync)
  • new and updated records (ongoing events)

What changed

This release expands virtual webhooks into a complete data delivery layer:

  • historical data delivered through webhooks
  • configurable event types (created, updated)
  • field-level payload control
  • configurable detection intervals
  • manual trigger support for testing
  • webhook observability (audit trail + health status)

1. Historical data via webhooks (initial sync)

You can now retrieve existing data through webhooks by setting:

include_all=true

Instead of:

  • calling list endpoints
  • managing pagination
  • storing checkpoints

Unified handles:

  • pagination
  • rate limits
  • retries and backoff

Your endpoint receives paged data until backfill completes, then transitions automatically to incremental updates.

2. Separate event types (created, updated)

Webhooks can now be scoped to specific event types.

This allows you to:

  • isolate insert vs update logic
  • simplify downstream processing
  • improve observability in logs and audit trails

3. Field-level payload control

You can define which fields are included in webhook payloads.

  • default: all fields (excluding raw)
  • optional: pass a subset via fields

This reduces payload size and unnecessary processing.

4. Configurable detection intervals

Virtual webhook intervals are configurable per connection:

  • free tier: ~60 minutes
  • paid plans: as frequent as ~1 minute

This allows you to balance:

  • freshness requirements
  • API usage
  • product experience

5. Manual trigger for testing

Webhooks can be triggered on demand.

This removes the need to wait for scheduled intervals when:

  • testing integrations
  • validating payloads
  • debugging delivery

6. Webhook observability (audit trail + health)

You now have visibility into webhook execution:

Audit trail

  • recent runs
  • records delivered
  • execution history

Health status

  • indicates delivery reliability
  • surfaces persistent failures
  • backed by retry logic and backoff handling

What this replaces

Before:

  • polling jobs for backfill
  • separate pipelines for historical vs incremental data
  • custom retry + pagination handling

After:

  • a single webhook subscription
  • one event stream
  • no polling infrastructure

Billing overview

Virtual webhooks are billed based on data retrieved and processed, not empty intervals.

Billable

  • successful data reads (including pagination)
  • dispatch attempts

Not billable

  • retriable read errors (e.g. rate limits)

This aligns usage with actual data movement, not polling frequency.

Availability

Webhooks are available across all Unified plans, with interval limits based on tier.

Key takeaway

Webhooks are no longer just for incremental updates.

With initial sync + ongoing delivery in one flow, they replace:

  • polling
  • backfill jobs
  • sync pipelines

Your integration layer becomes event-driven by default, even for APIs that don't support webhooks.

All articles