Cloud Flows: Dataverse Triggers & Expressions

Making flows that react intelligently

Dataverse trigger configuration

Trigger	Fires When	Key Settings
When a row is added	New record created	Table name, scope (org/BU/user)
When a row is added, modified or deleted	Any CRUD operation	Change type (added/modified/deleted), column filter, row filter
When an action is performed	Custom API or bound action called	Table name, action name

Trigger filters (critical for performance)

Without filters: The flow triggers on EVERY change to the table — even changes your flow does not care about. This wastes flow runs and can cause throttling.

With filters: The flow only triggers when specific conditions are met:

Filter Type	How to Configure	Example
Column filter	Select specific columns	Only trigger when statuscode or priority changes
Row filter	OData $filter expression	statuscode eq 1 — only trigger for Active records
Scope	Organisation, Business Unit, User	Organisation = all records; User = only records owned by the flow owner

OData filter examples

// Only trigger for high-priority orders
priority eq 1

// Only trigger for records in specific region
regioncode eq 'NZ' and statuscode eq 1

// Only trigger for records with value over $10,000
totalamount gt 10000

💡 Exam tip: Column filters prevent infinite loops

A common exam scenario: “A flow triggers when a record is modified. The flow updates a field on that same record. This creates an infinite loop.” The fix: use a column filter to trigger ONLY when specific fields change — not the field your flow updates.

Example: Flow triggers when statuscode changes. Flow updates processeddate. Since processeddate is not in the column filter, the update does not re-trigger the flow.

Complex expressions

Power Automate expressions use functions for data transformation:

String functions

// Concatenate
concat('Hello, ', triggerOutputs()?['body/firstname'], '!')

// Format a date
formatDateTime(triggerOutputs()?['body/createdon'], 'dd MMM yyyy')

// Extract substring
substring(variables('orderNumber'), 0, 3)

// Replace text
replace(body('Get_item')?['description'], 'OLD', 'NEW')

Conditional expressions

// If-then-else
if(equals(triggerOutputs()?['body/priority'], 1), 'Critical', 'Normal')

// Coalesce (first non-null value)
coalesce(triggerOutputs()?['body/preferredname'], triggerOutputs()?['body/firstname'])

// Null check
if(empty(triggerOutputs()?['body/email']), 'No email', triggerOutputs()?['body/email'])

Date/time functions

// Add days to a date
addDays(utcNow(), 7)

// Difference between dates
div(sub(ticks(variables('endDate')), ticks(variables('startDate'))), 864000000000)

// Start of month
startOfMonth(utcNow())

Retry policies

Actions in Power Automate can be configured with retry policies for transient failures:

Policy	Behaviour	Best For
Default	4 retries with exponential intervals	Most actions
Fixed interval	Retry at set intervals (e.g., every 30s)	When you know recovery time
Exponential	Increasing intervals (1s, 2s, 4s, 8s…)	API rate limiting
None	No retry — fail immediately	When retries would cause duplicates

Question

How do you prevent a cloud flow from creating an infinite loop when it modifies its own trigger table?

Click or press Enter to reveal answer

Answer

Use column filters on the trigger. Only trigger when specific fields change (e.g., 'statuscode'). The flow's own updates to other fields (e.g., 'processeddate') will not re-trigger the flow because those fields are not in the column filter.

Click to flip back

Question

What is the coalesce() expression in Power Automate?

Click or press Enter to reveal answer

Answer

coalesce() returns the first non-null value from a list of arguments. Example: coalesce(preferredName, firstName, 'Unknown') returns preferredName if it exists, otherwise firstName, otherwise 'Unknown'. Useful for handling optional fields with fallback values.

Click to flip back

Question

What retry policy should you use when calling an API that returns HTTP 429 (rate limit)?

Click or press Enter to reveal answer

Answer

Exponential backoff — increasing intervals between retries (1s, 2s, 4s, 8s...). This gives the API time to recover and reduces the chance of hitting the rate limit again. The Retry-After header from the 429 response should be respected when available.

Click to flip back

Knowledge Check

A cloud flow triggers when any record in the Contact table is modified. The flow updates the 'Last Processed' field on the same Contact. Users report that the flow runs hundreds of times per minute. What is the root cause and fix?

AThe flow has a bug — add a condition to check if 'Last Processed' is emptyBThe flow is in an infinite loop — the update triggers the flow again. Fix: add a column filter to trigger only when specific fields (not 'Last Processed') changeCThe Dataverse connector is misconfigured — switch to a different trigger typeDToo many users are editing contacts simultaneously — reduce the scope to 'User'

Check Answer

Next up: Cloud Flows: Security, Errors & Child Flows — managing sensitive data, error handling patterns, and reusable flow logic.