Overview
Business Activity Monitoring (BAM) supports two data collection approaches: the Push Model and the Pull Model. Both let you track business transactions and monitor outcomes in Turbo360, but they work differently and suit different integration architectures. Use this article to decide which model fits your environment before you begin setup.
Business value
Choosing the right model upfront avoids costly rework later. The decision affects how you instrument your integrations, what infrastructure you deploy, and how quickly tracked data appears in BAM. Understanding the trade-offs before you start means your team spends time building the right solution — not migrating from the wrong one.
How it works
Push Model
In the Push Model, your integration solution actively sends business events to Turbo360 via the BAM API. Each time a meaningful checkpoint occurs — an order received, a payment validated, a shipment dispatched — your solution emits an HTTP call to BAM with the relevant tracked property values.
BAM receives the event, matches it against the configured business process and transaction, and updates or creates the corresponding transaction instance in real time.
The Push Model uses the Custom Logging transaction type. Business processes and transactions must be configured with this type before events can be received.
How data flows:
- Your integration solution reaches a tracked stage.
- It calls the BAM API with an event payload.
- BAM matches the event to a business process and transaction.
- The transaction instance is created or updated immediately.
Supported technologies:
| Technology | Integration approach |
|---|---|
| Logic App Standard | Logic App Connector, APIM proxy, HTTP connector |
| Logic App Consumption | Logic App Connector, APIM proxy, HTTP connector |
| Azure Functions | NuGet package (C#), HTTP calls (other languages) |
| API Management | Policy fragment with HTTP request |
| Power Automate | Connector |
| Data Factory | Web Activity |
| Custom code | NuGet package (C#), HTTP calls (other languages) |
| PowerShell | HTTP call |
| MuleSoft, Boomi, Frends | HTTP call |
Pull Model
In the Pull Model, BAM queries existing data sources — Azure Application Insights, Log Analytics workspaces, or Azure Data Explorer (ADX) — on a schedule. You configure BAM to run queries against these sources at defined intervals, and BAM constructs transaction instances from the results.
No changes are required to your integration solution. BAM reads data that is already being written to your observability platform.
How data flows:
- Your integration solution runs and writes telemetry to App Insights, Log Analytics, or ADX as usual.
- BAM executes a scheduled query against the configured data source.
- Query results are mapped to business processes and transaction stages.
- Transaction instances are created or updated based on the query output.
Comparison
| Factor | Push Model | Pull Model |
|---|---|---|
| Data latency | Real time — events arrive as they occur | Near real time — data available after the next scheduled query |
| Integration changes required | Yes — you instrument your solution to emit events | No — BAM reads from existing data sources |
| Infrastructure to deploy | BAM environment (Storage Account, Azure Functions) | Service principal with read access to App Insights, Log Analytics, or ADX |
| Data source | BAM API | Azure Application Insights, Log Analytics, Azure Data Explorer (ADX) |
| Transaction type | Custom Logging | App Insights, Log Analytics, ADX |
| Control over tracked data | High — you define exactly what is captured at each stage | Medium — limited to what your telemetry already records |
| Suitable for new integrations | Yes | Yes |
| Suitable for existing integrations without code changes | No | Yes |
When to use each model
Choose the Push Model when:
- You are building a new integration and can instrument it from the start.
- Your business process requires real-time transaction visibility.
- You need to capture custom business properties not available in standard telemetry.
- You want precise control over what data is tracked at each stage.
Choose the Pull Model when:
- You have existing integrations that already write to App Insights, Log Analytics, or ADX.
- You cannot or do not want to modify your integration solution.
- Near-real-time visibility (query interval) is sufficient for your monitoring needs.
- You want to get started quickly without deploying additional infrastructure.
You can use both models within the same Turbo360 environment. Different business processes can use different models depending on the integration they monitor.