How to Trigger Power BI Refresh with Zapier or Make

How to Trigger Power BI Refresh with Zapier or Make

A fixed refresh schedule works until your data arrives at an unexpected time. A file lands late, a month-end close finishes early, or a CRM import completes ten minutes after the report has already refreshed. People then read yesterday’s numbers and assume they are current.

This guide shows you how to trigger Power BI refresh from Zapier or Make, and how to call the Power BI refresh API directly when you need more control. It also covers what can go wrong, because an automated refresh that fails silently is worse than a schedule everyone can see.

The item being refreshed is a semantic model (called a dataset in some tools). We use “semantic model” from here on.

Quick answer

• To trigger Power BI refresh, send a refresh request for the semantic model when your source data is ready, instead of waiting for a clock.

• Zapier and Make may offer a ready-made Power BI action or module for this. Check each tool’s current app list, because available actions change.

• If there is no suitable action, or you want more control, send an authenticated POST request to the Power BI refresh API from an HTTP or webhook step.

• Add three safeguards: avoid overlapping refreshes, check the result afterwards, and alert someone on failure.

• Refresh limits depend on your license and capacity, so confirm them before you automate.

When a trigger-based refresh beats a fixed schedule

Scheduled refresh is simple, and for many reports it is enough. A trigger-based refresh is worth the effort when timing matters more than convenience.

Good candidates include:

• Unpredictable arrival times. A partner file or manual upload lands at different times each day.

• Upstream dependencies. A warehouse load or overnight job must finish before the report refreshes.

• Business events. A deal closes in Zoho CRM or an invoice is marked paid in Zoho Books. These are examples of trigger sources, not a claim that any integration exists.

• Fixed review moments. A Monday meeting needs fresh data just before it starts.

Triggers are a poor fit when the source changes constantly or the semantic model takes a long time to refresh. Firing on every record change can queue up refreshes you don’t need and exhaust your daily allowance. A common middle path is to keep a schedule as a safety net and add a trigger for the moments that matter.

Option 1: Trigger Power BI refresh with the native Zapier action

Zapier has a Power BI app. Whether it includes an action that refreshes a semantic model (a “dataset” in Zapier’s wording) depends on the current version of the app, so check before you build. Open the Power BI app’s action list in the Zap editor and look for a refresh-related action.

If one exists, the setup looks like this:

1. Choose a trigger. Pick the event that tells you the data is ready, such as a new spreadsheet row, a status change in Zoho CRM, or a message from a data pipeline.

2. Add the Power BI action. Connect your Microsoft account when prompted, then select the workspace and semantic model.

3. Filter if needed. Use a filter step so only the right events refresh the model, for example when a status equals “Closed won”.

4. Test. Run the Zap and confirm a new entry appears in the semantic model’s refresh history in Power BI.

5. Turn it on.

If the Power BI app has no refresh action, or its options are too limited, use Option 3. Zapier’s webhook or custom-request features may be able to send the request.

The trade-off: native actions are quick and need no code, but they may not let you read the refresh result or set refresh options.

Option 2: Trigger Power BI refresh with the native Make module

Make also has a Power BI app. Module names change over time, so look for a module that refreshes a dataset or semantic model rather than relying on a name from this article.

If you find one:

1. Start the scenario with an instant trigger (such as a webhook) or a module that watches your source app.

2. Add the Power BI module. Create a connection, sign in, and select the workspace and semantic model.

3. Run once and confirm the refresh appears in Power BI’s refresh history.

4. Add error handling. Attach an error-handling route to the module that sends an alert if the step fails.

5. Activate the scenario.

If you can’t find a suitable module, use Make’s HTTP capability to call the API, as described next. Searching for “make power bi dataset refresh” will also turn up community threads, but treat them as hints and test against your own tenant.

Option 3: Call the Power BI refresh API directly

Calling the REST API gives you the most control. It works from Zapier (through a webhook or custom-request step), Make (through an HTTP module), or any tool that can send an authenticated HTTPS request. The Power BI refresh API takes a POST request to the semantic model’s refreshes path.

Example request:

POST https://api.powerbi.com/v1.0/myorg/groups/{workspaceId}/datasets/{semanticModelId}/refreshes

Authorization: Bearer {access token}

Content-Type: application/json

{

"notifyOption": "MailOnFailure"

}

Key points:

• The IDs come from the workspace and semantic model URLs in the Power BI service.

• A successful response means Power BI accepted the request, not that the refresh finished.

• To check the outcome, send a GET request to the same path and read the most recent entry in the refresh history. Its status tells you whether the refresh is still running, completed or failed.

• Advanced refresh options may exist for some capacity types, but they are outside this guide. Confirm what your license supports.

For authentication, request an access token from Microsoft. Access tokens are short-lived, so configure the tool to fetch a fresh token (or use its built-in OAuth connection) instead of pasting one into the step.

Making it reliable

Triggering a refresh is the easy part. These safeguards keep it dependable.

Avoid overlapping refreshes

If your trigger fires several times in a short period, such as ten rows added in a minute, you can send ten requests. At best the extras are wasted. At worst they fail or use up your daily allowance.

• Batch events by waiting a few minutes, then send one request.

• Check the latest refresh status first and skip the request if one is still running.

• Trigger from a single “all done” signal instead of every record.

What the service does when you request a refresh during a running one can change, so test it and record the result.

Check the result

Add a step that waits, then reads the refresh history and checks the latest status. In Option 3 this is a second request. Native actions or modules may or may not expose the status.

Alert on failure

Notify the report owner by email or chat when the status is failed, or when a refresh is still unfinished after longer than usual. Include the semantic model name and the time. Also use Power BI’s own failure notifications, and add error handling in Zapier or Make so a failed automation step is visible too.

Respect refresh limits

Daily refresh limits depend on your license and capacity type. Scheduled and triggered refreshes may count toward the same total, so count both before you automate. Check Microsoft’s current documentation for your setup.

Permissions and authentication basics

The account or app that sends the request needs permission on the semantic model, usually as its owner or through sufficient workspace rights. The semantic model’s own data source credentials must also be set up, because the refresh uses them.

Two common routes:

• A named user account. Easy to start, but it breaks when the password changes, multi-factor prompts appear, or the person leaves.

• A service principal (an app registration in Microsoft Entra ID). Better for long-term automation, but your Power BI admin must allow service principals to use Power BI APIs in tenant settings, and the app must be added to the workspace.

Whichever you choose, use a dedicated identity rather than a personal one, keep secrets in the tool’s secure connection storage, and grant the minimum access needed. Ask your Power BI admin early, because tenant settings are a common blocker.

Common mistakes

• Triggering on every record change instead of a batched or “finished” event.

• Treating an accepted request as a successful refresh.

• Using a personal account that later expires or leaves the company.

• Forgetting data source credentials, so the API call succeeds but the refresh fails.

• Pasting a short-lived access token into an HTTP step.

• Removing the schedule entirely, with no fallback.

• Sending alerts to a channel nobody reads.

• Skipping the test against refresh history before going live.

FAQ

Can I trigger a Power BI refresh without Premium?

Often yes for basic refresh requests, but daily limits are tighter on shared capacity, and some advanced options may need Premium or Fabric capacity. Confirm what your license allows.

Do I need Zapier or Make to use the Power BI refresh API?

No. Any tool or script that can send an authenticated HTTPS request can call it. Zapier and Make are convenient when you already use them to connect other apps.

How do I know the refresh worked?

Read the semantic model’s refresh history after the request, either in Power BI or through the API, and alert on any status other than completed.

Should a trigger replace my scheduled refresh?

Not necessarily. Many teams keep a schedule as a backup and add triggers for time-sensitive moments.

Need help automating Power BI refresh?

Cee Apps is a Power BI services partner. It can help in three areas:

• Business intelligence reporting: report and dashboard design in Power BI.

• ETL and data warehousing: preparing and organizing the data that feeds Power BI.

• Scripting and automation: Zapier, Zoho Flow, Make.com and Google Apps Script.

If you want to get our services on Power BI for your business, talk it through with the team.

Book a Meeting