Ever had a Power Automate flow fail and wondered, "Why didn’t it just try again?"
You're not alone. One of the more powerful - but often overlooked - features in Microsoft Power Automate is the retry policy. It’s the safety net that makes your flows more resilient when things don’t go exactly as planned.
In this post, we’ll break down what a retry policy is, how it works behind the scenes, when you should customize it, and how to avoid common pitfalls. Whether you’re a beginner or have a few flows under your belt, this guide will help you level up your automation game.
What Exactly Is a Retry Policy?
Simply put, a retry policy tells Power Automate how to respond when an action doesn’t succeed the first time - like when a service is temporarily down or a connection times out.
Instead of giving up right away, Power Automate can try again automatically. This is super helpful when you’re dealing with things like:
- Unstable internet connections
- APIs that hit rate limits
- Temporary server issues
The Default Retry Behavior
By default, many actions in Power Automate already have retry logic built in. If something goes wrong temporarily, the platform will retry the action up to 4 times using something called exponential backoff.
What’s exponential backoff?
It’s a fancy way of saying that the time between each retry keeps getting longer. So if the first retry happens after 5 seconds, the next might be 10, then 20, and so on.
This approach prevents hammering the system with constant requests - and gives it a chance to recover.
Want More Control? Customize It.
Sometimes, the default retry policy isn’t quite right for your scenario. Maybe the system you’re calling has specific retry rules, or maybe you don’t want it to retry at all.
You can adjust this easily:
How to Customize Retry Settings:
- Click on the action you want to change.
- Tap the three dots (...) > Settings.
-
Under Retry Policy,
choose:
- None – Don’t retry at all.
- Fixed Interval – Retry after the same delay each time.
- Exponential Interval – Retry with increasing delays.
- Set how many times to retry and how long to wait between tries.
1. "None" – No Second Chances
Choosing None means your flow won’t retry the action at all. If it fails once, it’s done. The flow moves on (or stops, depending on how you’ve set it up).
When should you use it?
Go with this if:
- The action must not run more than once - like sending a payment or submitting a form.
- You want the flow to fail fast and hit an error path quickly.
- Retrying would create duplicates or conflicts (like multiple emails).
Heads-up:
This policy is pretty unforgiving. Even a tiny glitch can cause your flow to stop. So, use it only when you're sure retries aren't safe.
2. Fixed Interval – Try Again After a Break
With Fixed Interval, Power Automate will retry the action a few times, waiting the same amount of time between each try.
Let’s say you choose a fixed interval of 10 seconds and 3 retries. That means Power Automate will try once, then again in 10 seconds, again in another 10, and a final time after another 10—then give up if it still fails.
You can customize:
- How many times it retries
- How long it waits between each try
3. Exponential Interval – Smart Retrying That Backs Off
Instead of retrying every 30 seconds like Fixed Interval, Exponential Interval increases the wait time after each attempt.
Example:
- First retry: 5 seconds
- Second retry: 10 seconds
- Third retry: 20 seconds
- …and so on
Perfect for:
- APIs that throttle or rate-limit requests
- Systems under heavy load
- General-purpose retrying when you're not sure how long a system might take to recover
Why it's awesome:
This is actually the default retry policy for most actions in Power Automate - and for good reason. It gives systems time to breathe and reduces the risk of overloading anything.
When Should You Use Custom Retry Policies?
Here are some common situations where tweaking the retry settings makes sense:
- Calling an external API that might temporarily throttle requests
- Working with cloud storage like SharePoint or OneDrive where things might lag
- Sending emails through servers that occasionally delay or reject requests
- Handling approvals where duplicate requests could cause issues (you might want no retries here)
Best Practices to Keep in Mind
If you’re customizing retry settings, here are a few tips to do it right:
- Start with exponential backoff – It’s the safest way to avoid overloading a system.
- Avoid retry loops – Don’t set retries too high or too frequent; it can make things worse.
- Use error handling – Combine retry policies with Scope controls and “Configure Run After” options.
- Log your retries – Use Compose or Append to Array to track what’s happening during retries.
- Know when to stop – If a retry won’t help (e.g., the request is invalid), disable it.
How to Know If a Retry Happened
Curious if a flow retried an action?
Check the flow run history:
- Open the flow run
- Click the action
- You’ll see retry attempts, delays, and error messages (if any)
This is great for troubleshooting or refining your logic.
Wrapping Up
Retry policies might not be the flashiest feature in Power Automate, but they’re absolutely essential for building dependable, real-world flows.
Whether you’re working with APIs, cloud storage, or approvals, adding (or disabling) retries at the right time can help you:
- Avoid flow failures
- Reduce false alerts
- Handle real-life network issues gracefully
Next time a flow action fails, don’t just ask “what went wrong”—check the retry settings. A small tweak there could save you a lot of time and headaches.
FAQs About Power Automate Retry Policies
Have a good day!