Skip to content
Tutorials Center
My learning

Topics

Lesson 1 of 38 ยท Getting Started

What Is Microsoft Power Automate? Flows, Triggers and Actions

Learn what Microsoft Power Automate does, how a flow is built from a trigger and actions, what the product family includes, and when to automate a task.

Microsoft Power Automate is a service for automating work between apps and services. You describe a rule such as "when an email with an attachment arrives, save the file to a folder and tell the team", and the service runs it for you every time, without anyone clicking anything. The rule is called a flow.

You build flows in a visual designer instead of writing a program. That makes Power Automate popular with office staff, analysts and administrators, and it is also used by professional developers who want to connect systems quickly. This lesson explains the vocabulary and the shape of every flow you will build in this course.

Power Automate is a cloud product from Microsoft, so it cannot run on this site. The lessons describe the menus, buttons and settings in words. Every TryIt block is a small Python stand-in, labelled as such, that shows the same idea (for example how a trigger passes data to an action). The stand-ins are not Power Automate.

The core idea: trigger, actions, result

Every flow has the same skeleton.

  1. A trigger is the event that starts the flow: a new email, a file created, a button press, a clock reaching 8:00.
  2. One or more actions are the steps that run after the trigger: send a message, create a row, copy a file, ask for approval.
  3. Data moves between steps as dynamic content: the sender of the email, the file name, the row id.

Anatomy of a flow

Trigger The event that starts it
Actions Steps that run in order
Dynamic content Data passed along

Every flow starts with exactly one trigger.

1 / 2

A flow never starts by itself without a trigger, and it has exactly one. It can have as many actions as you like, including conditions, loops and branches that you will learn in later chapters.

A flow as a simple model

Here is a stand-in written in Python. It is not Power Automate; it models a flow as a trigger payload plus a list of actions, so you can see how data moves.

Python Stand-in: a flow is a trigger plus actions

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Both actions read the same trigger output. In the real designer you pick those values from a list called Dynamic content instead of typing names.

The Power Platform family

Power Automate is one of several products that share data and licensing under the name Power Platform.

Product What it is for
Power Automate Automating processes (this course)
Power Apps Building forms and business apps
Power BI Reports and dashboards
Power Pages Public websites with data
Copilot Studio Chatbots and agents
Dataverse A database that stores data for all of them

Flows often live next to a Power Apps app: the app collects input, and a flow processes it.

Three kinds of automation

The word "flow" covers a few different tools.

  • Cloud flows run in Microsoft's cloud and connect services with APIs. They are the main subject of chapters 1 to 5 and 7 to 9.
  • Desktop flows (previously called UI flows) drive an application on a Windows computer by clicking and typing, like a robot. They are covered in chapter 6.
  • Business process flows guide a person through the stages of a process inside a Dataverse app. They are mentioned in chapter 8.

The next lesson goes into the types and the licensing that goes with them: Flow types and licensing.

When should you automate?

Automation is worth it when the task is repeated, rule based and annoying. Ask these questions:

  • Does it happen often, or does it block other people while they wait?
  • Can the steps be written as "if this, then that"?
  • Is the data available in a service that has a connector?
  • What happens if the automation is wrong? Can a person review the result?

A task that happens once a year, needs human judgement at every step, or involves systems with no connector is a poor first candidate.

The next stand-in scores a task idea with those questions. It is a plain Python helper, useful as a checklist.

Python Stand-in: is this task worth automating?

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Common mistakes

  • Starting with a huge process. Begin with a flow of three steps and extend it.
  • Forgetting that flows run as a person. Many connections use the account of whoever created them, which matters when that person leaves (see Connections and authentication).
  • Treating the designer as a toy. Flows have run history, versions, ownership and limits, like any other software.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What starts a flow?

    Every flow has exactly one trigger, which is the event that starts a run.
  2. 2. How do steps in a flow share data?

    Outputs of the trigger and earlier actions appear as dynamic content that later steps can use.
  3. 3. Which kind of flow controls an application on a Windows desktop by clicking and typing?

    Desktop flows automate user interfaces on a Windows machine, in the style of robotic process automation.
  4. 4. Which task is the best first candidate for automation?

    Frequent, rule-based work on systems that have connectors gains the most from automation.

Summary

  • A flow is a trigger followed by actions; data travels as dynamic content.
  • Power Automate belongs to the Power Platform with Power Apps, Power BI, Power Pages, Copilot Studio and Dataverse.
  • Cloud flows, desktop flows and business process flows solve different problems.
  • Automate frequent, rule-based tasks and keep a human review where mistakes are costly.

Next: Flow types and licensing