Thinking in Workflows: Triggers, Actions, Conditions

The Automation Alphabet

Every automation, from a simple email auto-reply to a complex multi-system data pipeline, is built from three building blocks:

1. Triggers – What starts the workflow
2. Actions – What the workflow does
3. Conditions – What decisions the workflow makes

That’s it. Master these three concepts and you can design any automation, in any tool, for any process. They’re the alphabet of automation – once you know the letters, you can write anything.

What You’ll Learn

By the end of this lesson, you’ll understand triggers, actions, and conditions deeply enough to break down any process into automatable components. You’ll also learn to map workflows visually before building them.

From Opportunity to Design

In Lesson 1, you identified automation candidates in your work. Now we’re going to learn the language for describing how those automations should work. Think of Lesson 1 as saying “I want to go to Paris.” This lesson is learning to read a map.

Triggers: What Starts Your Automation

A trigger is the event that says “go.” Every automation needs exactly one trigger (though some tools allow multiple trigger conditions).

Types of Triggers

Time-based triggers:

• At a specific time: “Every Monday at 9 AM”
• On a schedule: “Every 4 hours”
• On a date: “On the 15th of each month”

Event-based triggers:

• Something happens: “When a new email arrives”
• Something changes: “When a spreadsheet row is updated”
• Something is created: “When a new customer signs up”

Condition-based triggers:

• A threshold is crossed: “When inventory drops below 10 units”
• A state changes: “When a deal moves to ‘Closed Won’”
• A value matches: “When an email contains the word ‘urgent’”

Manual triggers:

• A button is pressed: “When I click ‘Run Report’”
• A form is submitted: “When someone fills out the intake form”

Examples:

Quick Check

Look at the automation candidates you listed in Lesson 1. For each one, identify the trigger. Is it time-based, event-based, or condition-based? If you can’t identify a clear trigger, the task might not be a good automation candidate.

Actions: What Your Automation Does

Actions are the actual work. Each step in your automation is an action. Actions fall into categories:

Communication actions:

• Send an email
• Post a Slack/Teams message
• Send an SMS notification
• Create a calendar event

Data actions:

• Create a new record
• Update an existing record
• Delete a record
• Move data from one place to another

Document actions:

• Create a new document
• Add content to a document
• Generate a PDF
• Upload a file

System actions:

• Call an API
• Run a script
• Wait for a specified time
• Log an event

Example workflow with multiple actions:

Trigger: New customer signs up

Actions:

1. Create customer record in CRM
2. Send welcome email to customer
3. Create onboarding task for customer success team
4. Add customer to email marketing sequence
5. Post notification in #new-customers Slack channel

Each action takes input (data from the trigger or a previous action) and produces output (data or a state change that subsequent actions can use).

Conditions: The Decision Points

Conditions are where your automation gets smart. They’re if/then branches that route the workflow based on data.

Simple conditions (if/then):

IF order value > $500
  THEN send to priority fulfillment
ELSE
  send to standard fulfillment

Multi-branch conditions:

IF customer is in North America
  THEN assign to NA support team
ELSE IF customer is in Europe
  THEN assign to EU support team
ELSE
  assign to global support team

Combined conditions (AND/OR):

IF email is from a VIP client AND contains "urgent"
  THEN route to manager immediately
ELSE IF email is from a VIP client
  THEN flag for priority response
ELSE
  route to standard queue

The Power of Conditions

Conditions transform simple automations into intelligent ones. Without conditions, your automation does the same thing every time. With conditions, it adapts.

Without conditions: “Every new lead gets the same welcome email.”

With conditions: “Enterprise leads get a personalized email from sales. Small business leads get a self-serve onboarding guide. Returning customers get a ‘welcome back’ message with their account history.”

Same trigger, three completely different experiences.

Mapping Workflows Visually

Before building any automation, map it on paper (or in a digital tool). Visual mapping reveals problems you can’t see in your head.

The Simple Map

For straightforward automations:

[Trigger] → [Action 1] → [Action 2] → [Action 3]

Example:

[New form submission] → [Create CRM record] → [Send welcome email] → [Notify sales team]

The Branching Map

For automations with conditions:

[Trigger] → [Condition] →  YES → [Action A]
                          →  NO  → [Action B]

Example:

[New order received] → [Order > $500?]
                         → YES → [Send to priority fulfillment]
                                  → [Assign VIP handler]
                                  → [Send premium confirmation]
                         → NO  → [Send to standard fulfillment]
                                  → [Send standard confirmation]

Using AI to Map Workflows

AI is excellent at helping you think through workflow logic:

I want to automate [process description].

Here's what I currently do manually:
1. [Step 1]
2. [Step 2]
3. [Step 3]
...

Please map this as an automation workflow with:
- The trigger (what starts this)
- Each action (what happens at each step)
- Any conditions/decisions (where the workflow branches)
- Data flowing between steps (what information each step needs)
- Potential failure points (where things could go wrong)

Format it as a clear, sequential workflow I can follow.

Quick Check

Take one of your automation candidates from Lesson 1. Try mapping it as a simple workflow: trigger, actions, and at least one condition. Can you see the complete path from start to finish?

Data Flow: The Invisible Thread

Here’s something most automation beginners miss: data flows through your automation like water through pipes. Each step needs input data, and each step can produce output data for the next step.

Example: New employee onboarding automation

Trigger: HR submits new hire form
Data available: name, email, role, start date, manager, department

Action 1: Create employee record in HRIS
Input: name, email, role, start date, department
Output: employee ID

Action 2: Send IT setup request
Input: name, email, role, employee ID, start date
Output: ticket number

Action 3: Send welcome email to new hire
Input: name, email, start date, manager name
Output: email sent confirmation

Action 4: Create onboarding checklist
Input: role, department, start date, manager, employee ID
Output: checklist URL

Action 5: Notify manager
Input: manager name/email, new hire name, start date, checklist URL
Output: notification confirmation

Notice how data from the trigger feeds into multiple actions, and some actions produce data (employee ID, ticket number, checklist URL) that later actions need.

Common data flow mistakes:

• Trying to use data that hasn’t been created yet (referencing a ticket number before the ticket is created)
• Not mapping where each piece of data comes from
• Assuming data will be in a certain format (dates, names, addresses vary)

Workflow Patterns

Most automations follow one of these patterns:

Sequential: Steps happen one after another

A → B → C → D

Branching: Different paths based on conditions

A → [Condition] → B or C

Parallel: Multiple actions happen simultaneously

A → B + C + D (all at once)

Loop: Actions repeat until a condition is met

A → B → [Done?] → No → B → [Done?] → Yes → C

Scheduled: Actions happen on a timer

[Every Monday] → A → B → C

Most real automations combine multiple patterns. A weekly report (scheduled) might pull data from multiple sources (parallel), check if there’s anything noteworthy (branching), and send different reports to different people (branching again).

Exercise: Design a Workflow Map

Choose one of these common scenarios (or use your own):

1. Client onboarding: When a new client signs a contract, automate the setup process
2. Content publishing: When a blog post is approved, distribute it across channels
3. Expense reporting: When an expense is submitted, route it through approval

For your chosen scenario:

1. Identify the trigger
2. List every action (at least 4 steps)
3. Add at least one condition/branch
4. Map the data flow (what data does each step need?)
5. Draw the complete workflow map

Use AI to help you think through steps you might miss:

I'm designing an automation for [scenario].
Here's my initial workflow map: [your map].
What steps or conditions am I missing?
What could go wrong at each step?

Key Takeaways

• Every automation is built from three components: triggers, actions, and conditions
• Triggers start the workflow (time-based, event-based, or condition-based)
• Actions do the work (communicate, manipulate data, create documents, integrate systems)
• Conditions make decisions (if/then branches that route the workflow)
• Always map workflows visually before building – it reveals gaps and edge cases
• Data flows through automations like water through pipes – track what each step needs and produces
• Most automations combine sequential, branching, parallel, and scheduled patterns

Next lesson: you’ll take one of your workflow maps and design a complete automation.