Build Flows

Introduction to Experience and Workflow Studio
Navigate Flows page in Workflow Studio

A Flow is a sequence of nodes that defines an automated workflow. A flow begins with a trigger node, which sets the process in motion based on specific events, and is followed by a series of action nodes that define the subsequent steps. The Flow Builder is the place where you can design your automated flows. You can configure each node individually, build in business logic, and connect to other data sources.

Navigating the Flow Builder interface

The menu bar at the top of the window allows you to add a flow name and flow description. Simply click on the text to edit it inline.

If your flow encounters issues, the menu bar alerts you with an error message indicator. Clicking this provides a breakdown of issues, opening the affected nodes for immediate action.

The menu bar also lets you know the save state of your flow. Flows save periodically after you make changes, and the status will either be 'Unsaved changes', 'Saving' or 'Saved'.

The Flow builder itself is divided into two sections:

• The toolbar (left): Here you can browse and select different nodes. You can click the arrow in the top right to minimise or maximise the toolbar.
• The canvas (right): Here you can drag and drop nodes and build your flow. The canvas lets you configure, rearrange, and build your flows. You can use the zoom slider in the top right to change the zoom scale, and the undo and redo controls to reverse changes.

At the bottom right, you will find button(s), applicable per the state that your flow is in.

• Exit
• Save draft
• Activate

What is a node?

All Pendula flows are made up of nodes. Nodes are modular and configurable. Each node represents a step in the workflow, and can be a process (sending or receiving data) or an activity (a touchpoint with a customer, like an SMS conversation). When nodes are put together, they create the customer experiences we call flows. Let's take a look at each node type:

Trigger nodes

These start the flow. They are event based (when something happens, start the flow) and are configurable.

Messaging nodes

These nodes facilitate communication through channels such as SMS, email, and social messaging supporting both outbound messages and two-way conversations.

Data nodes

These nodes enable API interactions, allowing for the transmission and receipt of data between the flow and external systems.

Workflow nodes

These nodes manage flow logic, allowing decision making based on predefined conditions and branching into multiple pathways based on distinct criteria.

Anatomy of a node

1. Guide Text: Positioned at the top, this text offers quick guidance or instructions specific to the node.
2. Node Body: The central part of the node that displays the Node Name and Node Description, which provide identification and details about the node's functionality.
3. Node Actions Panel: Located below the node body, this panel includes several icons for managing the node:
   • Settings: Opens the configuration modal for the node.
   • Duplicate: Allows copying of the node along with its settings.
   • Trash: Offers an option to delete the node, typically prompting a confirmation modal.
4. Outcomes: Indicated on the node, these direct the flow's path depending on the node's operation. Standard nodes may show a connector tab instead.
5. Drop zones: Highlighted areas on the canvas where nodes can be placed. These zones pulse to indicate where a node can be dropped.
6. Exit Node: Marks the path where the experience ends , guiding to subsequent steps or endings in the workflow.

Configuring nodes

Either click the settings icon (the cog) that appears on hovering over a node, or double-click the node itself to view and configure the settings. Different nodes have different configuration options.

Node settings are divided into two sections:

• The details pane (left): Here you can edit the name, description and guide text. These details are available on every node.
• The configuration pane (right): Here, you can adjust the node-specific settings, which vary depending on the type of node you are configuring.

Node names, descriptions and guide text don't affect anything to do with configuration.

Node management

• 
  
  
  • Find the toolbar on the left of the Flow Builder. This section houses all available nodes.
  • Use the search bar to find nodes by their name or description. The toolbar will dynamically filter to show nodes that match your search criteria.
  • Browse through categories to find the node you need
• 
  
  All flows need a trigger node to start the flow.
  • Click and drag a node from the toolbar onto the canvas. The drop zones will pulse where you can drop that particular type of node.
  • The node settings will open automatically after placing the node.
  • Once configured, click Save to apply your settings.

    While these node changes have been applied, the flow itself is saved periodically. You can also save the flow manually if you click Save draft in the footer.

    Nodes with multiple outcomes, such as the conversation node, will show helper paths on the canvas.

• 
  
  
  • Click and drag another node from the toolbar onto the canvas.
  • The drop zones will pulse where you can drop that particular type of node.
• 
  
  
  • Hover your mouse over the node you wish to delete.
  • Click on the bin icon that appears beneath it.
  When you delete a node with multiple outcomes, you can either choose to keep an outcome path, or delete all successive paths.
• 
  
  
  • Hover your mouse over the node you wish to duplicate.
  • Click on the copy icon that appears beneath it. Note that this is not supported on all nodes. If the icon is not visible, the node cannot be copied.

  The copied node will have "(Copy)" appended to the end of its name, and it will appear before the original node.

• 
  
  
  • To connect nodes and establish paths, drag from an outcome or the helper attached to the node if no outcomes are present.
  • Drag the connector to the desired node, and release to create a new path.

  You cannot remove a joined node path and orphan the child node; to remove a path you must delete the child node.

• 
  
  
  • Click and hold on a node to 'pick' it up. You'll notice you're now dragging an icon of the node.
  • Drop zones will pulsate indicating where you can place the node. Drag, and drop on top of a helper zone to place it.
• 
  
  
  • In addition to the drop zones that appear when moving a node, you'll notice that other nodes on the canvas will pulsate in the same way. This means you can swap places with them!
  • When you drag on top of one of these nodes, you'll notice two options, Swap and Replace. When hovering over the Swap side of the node, you'll notice it becomes highlighted. Simply drop the node here to swap them.

  Ensure your merge fields are still valid by checking the merge field data explorer on the newly swapped nodes.

• 
  
  
  • Drag a node on to the Replace side of the node, and drop the dragged node.
  • You will be prompted to confirm that you wish to replace the node configuration on the canvas with the one you're replacing it with.

  Ensure your merge fields are still available by checking the merge field data explorer on the newly swapped nodes.

• 
  
  Errors within nodes can be detected through two primary indicators:
  
  Within the Node
  If a node encounters an error, a message stating This node has some errors will prominently display within the nodule's interface. This alert serves as a direct indication that the particular node requires your attention for troubleshooting.
  
  Global Error View on Menu Bar
  Additionally, the Flow Builder provides a holistic view of errors across all nodes in a flow via the menu bar. If errors are present, an alert will list them collectively. Each error message is designed to guide you to the problem area within the node.
  

  For a detailed guide on understanding errors in Pendula, see Flow errors.

Introduction to Experience and Workflow Studio
Complete overview of the Flow Builder

A Flow is a sequence of nodes designed to execute an automated workflow.

An Experience is a unique journey of an individual recipient through a flow. Experiences are personalised paths shaped by the recipient's specific interactions and responses as they navigate through the nodes of a flow. This ensures that each recipient's engagement is tailored to their actions.

A Trigger is a type of node, defining the conditions that will trigger experiences for the flow. Triggers respond to defined events, such as receiving an inbound message or detecting a change in a CRM system, thereby activating the flow and commencing the sequence of actions.

Event-based triggers

These triggers await events such as incoming messages or system updates.

Inbound Message Trigger

Initiates a flow when a specific SMS message is received. It is configured to recognise predefined keywords or phrases within the text message, facilitating instantaneous engagement with customers.

For a detailed guide, see Inbound Message Trigger Node

Unmatched Message Trigger

Ensures that messages which do not match any active flows, are not lost. This trigger catches and handles all inbound messages that haven't been matched by other active triggers or conversation actions, maintaining a comprehensive customer communication loop.

For a detailed guide, see Unmatched Message Trigger

Webhook Trigger

Listens for data payloads sent from external systems to a specified URL. This trigger is essential for creating responsive workflows that integrate with a variety of external applications or services.

For a detailed guide, see Webhook Trigger

Amazon Kinesis Trigger

Starts a flow based on data received from an Amazon Kinesis stream. Amazon Kinesis is a scalable service from AWS that enables real-time processing of streaming large-scale data, which can be anything from log files to market data. When new data is detected in the stream specified for the trigger, it sets the flow into motion within Pendula.

For a detailed guide, see Amazon Kinesis Trigger

Advanced Integration

Pendula goes beyond basic integrations with a comprehensive Integrations Hub. For scenarios where an out-of-the-box integration is not available, the Studio allows Pendula users to create one-off custom integrations including systems like Salesforce, Hubspot, Zendesk and more.

For access to this feature, please reach out to our sales team who can provide more details and assist with enabling it for your account.

Setting up an event-based trigger

1. Begin by selecting the most suitable event-based trigger within the Flow Builder. Depending on your flow's requirements, you might opt for the Inbound SMS Trigger, Webhook Trigger, or Unmatched Message Trigger.
2. Define the specific conditions that will activate your trigger.
   • For the Inbound SMS Trigger, specify the exact words or phrases within the SMS body that should initiate the flow. This involves creating rules or recipes that can identify specific keywords, phrases, or formatted text such as email addresses or mobile numbers within an incoming message. Each path in the trigger can include multiple rules or recipes. When a reply matches more than one rule, the highest priority path is triggered first. This functionality ensures that every matching inbound message correctly initiates the intended workflow.
   • For the Webhook Trigger, the setup involves configuring the external system to interact with Pendula. This includes supplying the "POST URL" from Pendula, which is crucial for establishing a connection between the external application and the Pendula flow. Additionally, secure this connection by integrating the API key for authentication. While defining the exact JSON payload structure is optional, ensuring these elements are correctly configured is essential for triggering the workflow based on real-time data received from external applications.
   • The Unmatched Message Trigger acts as a catch-all for messages that do not match other triggers, so no explicit conditions are required for it.
3. Once you have defined the conditions, integrate your trigger with the subsequent steps in your flow. Each trigger should be connected to appropriate actions to ensure a cohesive and responsive experience.

All flows require a trigger node. You cannot have more than one trigger node.

You can replace a trigger node but you can’t duplicate or move one.

Deleting the trigger node will delete all nodes in the flow. 

Introduction to Experience and Workflow Studio
Complete overview of the Flow Builder

Actions are nodesInsert Definition Here that execute the subsequent steps after a triggersInsert Definition Here. These are nodes that execute subsequent steps, guiding the customer through a journey or handling data processes. Actions in the Flow Builder can be broadly categorised into messaging, data, and workflow nodes.

Messaging nodes

These nodes are responsible for sending communications to recipients. Each messaging node can be configured to send a specific message and, where applicable, respond based on customer interactions.

Outbound SMS

Sends a single SMS to a designated recipient. The node offers two outcome paths based on message delivery status—sent and not sent. Handle fallbacks and failures by tailoring experiences based on these outcomes.

Conversation

Handles two-way SMS interactions, enabling conversations that can adapt based on customer responses. Pendula’s industry-leading intent matching capability provides the user the ability to listen for any customer reply based on flexible rules, ranging from simple keywords to complex pre-built logic.

Outbound Email

Sends emails, which can be personalised and scheduled as part of the flow. Emails can be either built from scratch using email design pane or from importing HTML code.

Data nodes

Data nodes handle the flow's interactions with external data sources. This includes sending data to and receiving data from external APIs, allowing the flow to integrate seamlessly with other systems. These nodes are crucial for workflows that rely on dynamic data exchange.

Make a web request

Interacts with external services and APIs, allowing the flow to send data to or receive data from outside sources. It can be used to update a CRM record, log an event, or enrich the experience with external data.

Read more about integration-specific data nodes in the Integrations Hub.

Workflow nodes

Workflow nodes introduce logic and decision-making capabilities into the flow. These nodes ensure that the flow adapts and responds to varying inputs and scenarios.

Criteria Filter

Filters customers or data points based on predefined criteria, ensuring that only the relevant individuals or data proceed to the next step.

Criteria Split

Diverges the flow into different paths, each tailored to a specific customer segment or scenario.

Reporting Point

Marks a specific point in the workflow for reporting purposes, allowing for the collection of data on the flow’s performance.

Setting up an action

1. Choose the appropriate action node from the Flow BuilderInsert Definition Here based on what needs to be accomplished in the subsequent steps following a trigger.
2. Configure the node with the necessary details, whether it’s crafting a message template, defining API request parameters, setting criteria for filtering or splitting, or establishing timing for delays or reporting.
3. Use any subsequent nodes to ensure a seamless flow of the automated workflow.

Action nodes execute in a predefined sequence, as determined during in the flow. This sequential execution allows for complex workflows that can adapt based on dynamic inputs and conditions. Action nodes offer the capability to integrate with various systems and services, enhancing the flow's utility and adaptability.

Complete overview of the Flow Builder

Merge fields are a powerful tool to personalise your communications and customer journeys, using either the customer data from your systems of record, from external systems (weather services, traffic services etc.) or from the outputs of a node (for example, the recipient’s mobile number from the Inbound message trigger node). In this article, you’ll learn about merge fields and ways you can use them to personalise messages and customer journeys.

What are merge fields?

Merge fields are dynamic values that are replaced with the recipient-specific value when a flow goes live. The recipient-specific value is determined by the data brought into the flow. A merge field can be identified through its bold appearance, like this: firstName.

For instance, to personalise a message using the customer's first name, your message could use the merge field: firstName. When a flow is live, the merge field references all the data stored against the corresponding firstName field in your systems, which is retrieved by a prior node in your flow. The recipient would receive a message with the merge field replaced with their name:

Using merge fields in flows

You can view merge fields available to reference via the merge field pop up, or through the data explorer in the merge field tab. You can either search, or traverse prior nodes to find the merge field you want to reference. Once found, simply click to either insert (in the merge field pop up) or copy to clipboard (in the merge field tab).

Important considerations when viewing available merge fields:
Each node will only display merge fields generated by previous nodes along the same branch. You cannot reference merge fields from a node on a separate branch.
You can only select merge fields from nodes if they have outputs and are on the 'common path' on the current node.

Merge field pop up

The merge field pop up is a handy way to search and insert merge fields straight into an input field. It also easily lets you know whether you can use a merge field in that particular field.

How to use the merge field pop up

1. Click the merge field button. A search bar will appear.
2. Type a value in the search bar. The value will only be searched against the path of the merge field itself, not the description nor example.
3. Click the merge field you want to use to insert it, and voilà!

Merge field tab: Quick search

The merge field pop up is a handy way to search and insert merge fields straight into an input field. It also easily lets you know whether you can use a merge field in that particular field.

How to use the search bar

1. Type a value in the search bar. The value will only be searched against the merge field itself, not the description nor example.
2. Click to copy the merge field to your clipboard, and paste in the desired field.

Merge field tab: Explorer

The merge field explorer displays merge fields that can be referenced from nodes previous to this node.

How to use the merge field explorer

1. In the first column, click on a node item to view the merge fields within. This column displays all the nodes you can copy merge fields from, in respect to this node's position in the flow.
2. Click through the levels of data to select from, as indicated by a forward arrow icon. The merge field is indicated by a plus icon.
3. Click to copy the merge field to your clipboard, and paste into desired field within the node's configuration settings.

Further applications & considerations

Merge fields are not limited to message bodies. They can also be used in other fields across the Flow BuilderInsert Definition Here, such as criteria split fields and recipient fields.

Merge fields affect character count

The character count in the message body will not accurately reflect the rendered version of the merge fields.  For example, firstName can render as Lam (3 characters) or Maddie (6 characters).

Using merge fields with arrays

When using merge fields with arrays, Pendula currently defaults to rendering the first item in the array retrieved.

Use merge fields with arrays with caution, as you can't guarantee the array index exists in the database.

Using merge fields with a blank or non-existent value

If a merge field value is blank or non-existent in your data extension, it will simply render blank in the message to the recipient.

The exception to this is when you use merge fields for the sender email address or phone number. If the data for those fields does not exist, the message will not be sent.

Conditional logic & loops (Advanced)

Merge fields operates on the simple templating language called Handlebars. This templating language uses expressions, wrapped in double curly braces, to access values passed into the template. It supports conditional logic, loops, and helpers, enabling you to display specific blocks of information based on defined criteria.

For example, Betty, who had a timestamp against her visit, would receive a different message to Cody, who didn't.

If-else {{#if}}

If-else statements let you conditionally render blocks based on if-else logic arguments; if this, then do this. The {{#if}} block argument checks whether a value exists in the block, as opposed to a return of false, undefined, null, “”, 0 or {}.

Hey {{firstName}},
Thanks for your recent visit at Acme
{{#if webhookTrigger.date}}
on {{webhookTrigger.date}}
{{/if}}.
How did we go?

You can also use {{else}} inside of {{#if}} statements to determine what block will render if the initial{{#if}}block returns a false value.

Hey {{firstName}},
Thanks for your recent visit at Acme
{{#if webhookTrigger.date}}
on {{webhookTrigger.date}}
{{else}}
with our team
{{/if}}.
How did we go?

The example above would render differently if the date stamp was recorded or not.

Unless {{#unless}}

You can use the unlesshelper as the opposite of the if helper. Its block will likewise be rendered if the expression returns a false, undefined, null, “”, 0 or {}.

{{#unless hasActiveSubscription}}
Hey {{firstName}},
Remember, your subscription has expired. Renew now to continue enjoying our services!
{{/unless}}

For instances where more complex conditional logic is required, see Criteria Split Node.

For more information on using conditional logic with Handlebars, see here.

Segmenting audience using Workflow nodes
Criteria Filter and Criteria Split Workflow Actions

Configuring nodes like Criteria Filter and Criteria Split involves selecting from a range of data types. This article explains these types—formulas, fields, booleans, strings, numbers, arrays, and their application within node configurations.

Configuring criteria conditions

Criteria FilterInsert Definition Here and Criteria SplitInsert Definition Here nodes allow you to configure criteria using various data types. Field or Formula can be selected to define the source of the value for comparison, and an appropriate Value is then assigned. An Operator determines the relational condition, such as 'less than', 'equals', or 'contains'. The type on the right-hand side—be it Field, Formula, String, Number, Boolean, or their array variants—determines the nature of the comparison parameter, complemented by the corresponding Value input for the condition.

Formula

Formulas allow you to calculate expressions in real-time, when criteria are being evaluated during an experience.Insert Definition Here

Formulas can be entered on the left or right hand side of an operator when the formula type is selected. When the formula Value field is selected, a formula picker dropdown will appear to help you write formula expressions:

The following formulas and arguments are currently supported:

DateAdd("interval", number, instant)

Returns the result of adding the number of intervals to the given instant (e.g. add 5 days from now). An instant refers to a specific point in time that serves as a reference for calculating timestamps. This could be the current date and time or the start of an experience.

• interval — SECONDS, MINUTES, HOURS, DAYS
• number — The number of intervals to add (e.g. 5). This can be negative.
• instant — The instant to add to, representing a precise moment in time. This can be the current datetime returned by Now() or the starting point of an experience returned by ExperienceStart().

Now()

Returns the current datetime instant

To add 5 days to the current date and time, you would use DateAdd("DAYS", 5, NOW()). This formula calculates a new datetime that is exactly five days from the exact moment the formula is evaluated.

ExperienceStart()

Returns the instant the experience started (when the flow triggers for an individual recipient)

If you need to add 10 days to the moment an experience began, such as a user starting a trial or subscription, you would use DateAdd("DAYS", 10, ExperienceStart()). This helps in scenarios where actions or reminders need to be scheduled based on the start of an experience.

Examples

Field

The Field type allows referencing data directly from fields available in the flow. Begin typing in the input to search and select from available fields, which dynamically populate based on what's available in the merge field explorer. Field types also support referencing array values, such as string[], number[], or boolean[] where available.

For example, using the field City from your system of record, you could use the field type and add a value (for instance, Sydney) to filter recipients based on whether their city is Sydney or not. If your merge field explorer includes a field like Tags (a string[]), you can use the contains operator to check if a recipient has a specific tag.

String

The String type is used for text comparisons. Each character in the text is compared one at a time. You may also encounter string[] (array of strings) in payloads with multi-valued fields.

For example, you can specify exact text to match against a field's value, or use contains to find a substring within a string or a value within an array. The City field might be paired with the string value Sydney to target users specifically in that city, or Tags (as a string[]) to find anyone with the "VIP" tag.

When using operators like "greater than" or "less than" in string comparisons, the order may not be what you expect. For example, "2" is greater than "11", and "Cat" is less than "catalog" because uppercase letters sort before lowercase.

Boolean

The Boolean type supports true or false values and can also support boolean[] (array of booleans) in advanced cases. This type is ideal for fields that store binary data, such as isActive or emailVerified. Arrays can be used with the contains operator to check if any value in a list matches your criteria.

Number

The Number type is used for numerical comparisons, supporting both single values and arrays (number[]). This is useful for fields that store numeric values, such as age, quantity, or thresholds, and now for multi-valued fields such as IDs, codes, or amounts. Use the contains operator to check if a specific number appears in an array of numbers.

Usage scenario: Offer expiry example

As an example, you can use Criteria filter to create an offer expiry period in the flow.

When a recipient replies to the original outbound SMS, the reply will come back into Pendula, and the split will check whether it was received before the expiry - i.e. was it received 2 days before or 2 days after the original message was sent?

In the criteria builder:

1. LHS Select type = formula

2. Enter the formula = DateAdd ("DAYS", 2, ExperienceStart())

   (2 days from when the experience started)

3. Select operator = less than
4. RHS Select type = formula
5. Enter the formula Now()

This checks that the inbound message was received less than 2 days ago and the recipient has replied, so they can continue through the ‘success’ path of the split.

Working with Actions in Workflow Studio

Segmentation enables targeting specific subsets of an audience at any point within a communication flow. Pendula supports this through Criteria Filter and Criteria Split nodes, which segment recipients based on real-time data.

Criteria nodes segment recipients by evaluating real-time data. This segmentation enhances the effectiveness of communications by aligning them with the recipients’ current status or recent interactions. Creative and intelligent use of segmentation with multi-channel workflow allows you to seamlessly nurture your audience from initial engagement to brand advocate.

Criteria-based segmentation examples

Segment recipients within a flow by combining filters based on behavioural, demographic, social or custom attributes within your system of customer data. Any data you can send to Pendula, you can use to create tailored and unique experiences.

• Use criteria nodes to ensure communications are sent only to recipients who have not opted out. For instance, verify whether the 'optedOut' attribute in a CRM record is 'false' to filter and continue the flow with these recipients.
• Employ criteria to target recipients based on their recent engagement. By establishing a criterion to identify if 'lastCommunication' occurred less than five days prior, you can effectively engage active users with timely follow-ups or incentives.
• Segment recipients by their channel preferences to ensure the message is delivered (and read) on their favourite channel.
• Complex audience building with nested groups in criteria nodes.
• Use multiple segments throughout a flow to enhance the personalisation of your campaigns; Pendula’s Flow Builder supports the integration of multiple criteria splits, allowing you to tailor the journey based on customer replies or generative AI inputs.

Using advanced criteria nodes

Pendula enhances recipient segmentation through two workflow nodes:

Criteria Filter

This node assesses each recipient against predefined criteria. Recipients meeting the criteria continue on the "pass" path, while others are diverted to the "fail" path, streamlining the segmentation process for binary decision-making.

For a detailed guide on configuration, see Criteria Filter Workflow Action

Criteria Split

This node facilitates complex decision-making by directing recipients along multiple conditional paths. Recipients are evaluated sequentially against each path's criteria; the first matching path is chosen, or a default "Other" path if no criteria are met.

For a detailed guide on configuration, see Criteria Split Workflow Action

Segmenting audience using Workflow nodes

Criteria Filter

Criteria filter checks incoming data against the provided criteria. If the data matches your criteria, the experience continues along the success path. If the data does not match your criteria, the experience continues along the fail path.

Usage Scenario

The example below demonstrates the use of a Criteria Filter in an A/B testing scenario, initiated by a form submission. The Criteria Filter divides customers into two groups based on predefined criteria. Customers who meet the criteria for Group A receive Email A. Conversely, those who do not meet the criteria for Group A and fall into Group B receive Email B. This setup allows for direct comparison between different marketing approaches.

Criteria Split

Criteria split enables you to build multiple 'criteria paths', and checks recipients against the criteria for each path. If the criteria match, the data follows the corresponding path. If multiple path's match, the highest priority (top) path is followed. If no paths are matched, they follow the 'Other' path.

Usage Scenario

The example below showcases a Criteria Split in action, with a flow based on student project preferences from a form submission. The Criteria Split node examines the student's choice and directs the flow accordingly. Successful project matches trigger an email with relevant project materials, while mismatches result in an email informing the student of the issue and suggesting alternative actions.

Configuration of the Action

Groups and Rules

• Groups serve as containers for sets of rules, similar to how parentheses are used in mathematical equations to group together different parts of an expression. Choose Add a group to begin a new set of conditions that act as a separate logical statement within the criteria.
• Rules are the individual conditions that reside within a group. A single group may contain multiple rules connected by a shared logical operator, AND or OR. A rule can exist on its own without a group. Utilise the Add a rule option to introduce new conditions within an existing group.

• The option between AND and OR within the interface allows you to define how groups and rules interact.

  Choosing OR means that your segment will include recipients satisfying any one, some, or all of the rules within that group. For example, if your segment's rules are that a recipient’s preferred channel is Email OR SMS, then all recipients with either preferred channels will be included.

  Conversely, selecting AND ensures that only recipients who meet all conditions are included in your segment. For instance, if your segment's rules are that a recipient’s preferred channel must be SMS AND must have bought a subscription within the last 30 days, then only recipients who meet both criteria will be included.

  Additionally, this logic can be layered to create complex segmentations, such as segmenting recipients who satisfy a specific rule AND at least one of two other conditions. For example, a segment might include recipients who have clicked on a campaign link AND prefer Email OR SMS.

Operators and Types

• Operators define how a field’s value is compared to your criteria. Available operators include:
  • equals – The value must be an exact, case-sensitive match.
  • greater than / less than / less than or equal to – Numeric comparisons only.
  • null – The field exists but its value is empty (e.g. { "Age": null }).

  • contains – Checks if a value contains a substring, or if an array includes a value of the same type. This comparison is case-sensitive.
    Use contains to: 

    • Find a substring within a string (e.g. "Smith" in "John Smith").
    • Find a value inside an array (2 in [1, 2, 3]; "a" in ["a", "b", "c"]).

    Only valid comparisons are allowed; for example, you can’t check if a number (with type number) is contained within a string. 

• Types refer to the kind of data used in rules:

  • String – Text values.
  • Number – Numeric values.
  • Boolean – True or false.
  • Field – References another field from your system, similar to merge fields.
  • Formula – A calculated value based on other data.

  Choose the correct type for both sides of your comparison. Some operators (such as contains) support specific combinations only, such as string[] with string, or number[] with number.

Ensure the value you enter matches the selected type. For example, if comparing an array, the operator and comparison value must be compatible (e.g., a string with a string array). The system will warn you if you select an invalid combination.

For a detailed guide on types, see Working with Types: Formula, Fields and Values

Paths

This is only applicable for Criteria Split.

• In Criteria Split, each path can be given a path label, reflecting its unique conditions. These names appear on the canvas, offering a clear map of the flow’s branching logic. To set up multiple paths, label each with a relevant name that describes the segment or condition it represents.
• The paths within a Criteria Split are prioritised from top to bottom. Recipients are evaluated against each path’s criteria sequentially. If a recipient matches multiple paths, the system prioritises the topmost path. There’s also a provision for a fallback path, labeled typically as 'Other', which captures any recipients not meeting the criteria of the prior paths.
• Within the node, you can duplicate a path as well as change the order of path to set precedence.

Complete overview of the Flow Builder

An error is something that will prevent a nodeInsert definition here or a flowInsert definition here from executing successfully. Examples include misconfigured connections or merge fields that reference missing data.

Errors within nodes can be detected through two primary indicators:

• Within the Node If a node encounters an error, a message stating This node has some errors will prominently display within the node settings through a red outline and exclamation symbol. This alert serves as a direct indication that the particular node requires your attention for troubleshooting.
• Global error view on Menu bar Additionally, the Flow BuilderInsert definition here provides a holistic view of errors across all nodes in a flow via the menu bar. If errors are present, an alert will list them collectively.

Each error message is designed to guide you to the problem area within the node.

Working with Merge Fields in Workflow Studio

This article delves into how to detect and resolve merge field errors and warnings within your flows. When configuring flows, you might encounter various errors and warnings related to merge fields. Errors, depicted by a red outline and exclamation symbol on the node, must be resolved to activate the flow. Warnings, indicated by orange guide text and symbols, suggest potential issues that could lead to operational failures but do not prevent flow activation. Recognising and addressing these alerts is crucial for the seamless execution of your flows.

What are 'Errors' and 'Warnings'

Before diving into specific error and warning messages, it's important to differentiate between the two:

• Errors: These are critical issues that will definitely cause failures on flow activation. Therefore, errors in your flow will prevent activation and must be resolved in order to allow activation.
• Warnings: Warnings, on the other hand, are cautionary. They indicate potential issues that may cause failures, but do not prevent flow activation. Warnings can be dismissed, but the risk something might go wrong will remain. Warnings should be addressed based on the potential impact.

Common merge field errors and warnings

Below is a table outlining common merge field errors and warnings you might encounter, along with explanations to help you understand and resolve them.

• Outbound emails: Fully designed, beautiful emails made with our drag-and-drop builder
• Conversation nodes: Pre-written outbound SMS messages with configured reply paths
• Web requests: Pre-configured requests ready to send data to an endpoint

Adding to saved nodes can significantly speed up your workflow, allowing you to focus on creativity.

How do I add a node to saved nodes?

On the canvas, hover over the node you want to save, and click the grid with the plus icon that appears beneath it (between the duplicate and help icons).

This action will add the node to your saved nodes. Don’t worry—you can still edit the node on the canvas without affecting the saved version in your toolbar.

Where do I find my saved nodes?

In your toolbar, you'll see a 'Saved' tab next to the 'Nodes' tab. This is where your saved nodes will appear. Initially, it will be empty.

How do I remove a saved node?

You may quickly fill up your 'Saved' nodes tab, which is a good sign!

To clean it up, simply hover over a saved node in the toolbar, click the bin icon, and confirm deletion.

Deleting a saved node will not affect any existing flows—it will only remove it from the toolbar.

For a detailed guide on creating an authenticated web request credential, see Make a web request Authentication

Creating and configuring a web request

1. When building your flow, drag and drop the Make a web request node onto the canvas at the desired point where you want to make a web request.

2. In the Request details section, select the HTTP request method you want to use (GET , PUT , PATCH or POST ) The external service you are calling out to will inform you on what method is suitable.

3. Enter the API endpoint URL – this is the address where you want to send the request to.

   You can dynamically customise your request's URL with data from previous nodes in the flow by using merge fields directly into the URL parameters. This method allows the API endpoint to adapt based on context, effectively using real-time data to query specific information or actions relevant to the user.

   https://api.example.com/users?userid=userId&mobile=mobile
   

   In this configuration, userId and mobile are merge fields within the URL. They are dynamically replaced with actual values from previous nodes when the flow executes.

   When using the Make a Web Request action, Pendula automatically encodes URLs based on standard percent-encoding rules. In some cases, APIs that do not follow the URL encoding standard cannot be used with the Make a Web Request action.
   If your API requires a specific encoding format and you're seeing unexpected behaviour, please reach out to our support team with details.

4. Under Authentication, specify if your request requires verification. Select 'Bearer', 'Username & Password', or 'Headers' for authenticated access.

   If the credential hasn't been added for authenticated access, you can add one in the Integrations page. See Make a web request Authentication

5. You can also add headers in the Request headers section, if necessary.

   For headers related to security, such as 'Authorization' , add these through your web request credentials in the Integrations page. This helps centralise security management and simplifies the configuration process.

Configuring the request body

In the Request body section, you can select content type (JSON by default) and configure the request body payload using the property builder.

1. In the Property field, type the name of the data attribute you want to include in your request (e.g., username, email).
2. Select the Type from the dropdown: string, boolean, or number, depending on the data you need to input.
3. In the Value field, input the actual data or select the merge field icon to dynamically insert data from previous nodes.

If you wish to build your own JSON payload using code, you can select the ellipses next to the property builder, and choose Convert to code. Note that this action is not reversible, and you'll need to Reset to get the property builder back.

If you wish to send an empty request body for PUT , PATCH or POST actions, you can select the ellipsis icon > Convert to code.

Handling errors and retries in web requests

When using the Make a Web Request action, Pendula does not automatically retry requests for operations that modify data, such as PUT or POST requests. This is to prevent unintended duplicate actions in external systems—for example, avoiding multiple record creations if an API returns an error after partially processing a request.

If a request fails, the experience will follow the configured outcome path, allowing you to handle errors within your workflow. If you need to retry a request, you can design your flow to detect specific response codes (such as 500 Internal Server Error) and perform another web request action, or your preferred logic.

Since every external system behaves differently, we recommend reviewing how your target API handles errors and deciding whether retries are appropriate for your use case.

Advanced request preferences

• Follow Redirects: Allow requests to follow HTTP 3xx requests for a maximum of 10 redirects. This is useful for accessing requested resources that have moved to a new URL.

  • Retain the authorization header when a redirect happens to a different hostname.
  • Retain the original HTTP method after a redirect.

• Timeout: Wait for 100 seconds (configurable up to 360 seconds) for a response from the server before timing out. Adjust this setting if you require a longer or shorter wait time depending on the service you are interacting with.

These advanced settings ensure your web requests are handled efficiently and securely, providing greater control over how HTTP requests are processed within your workflows.

Outcomes

This example shows a successful response from the OpenAI API, where the model returns a generated response with a 2xx status code.

• outcome: success
• outcomeReason: ok
• response:
  • id: "chatcmpl-7Gk2BjG2D4R9U" (unique identifier for the response)
  • object: "chat.completion"
  • created: 1684176874
  • choices:
    • text: "The weather is nice in San Francisco."
    • index: 0
    • logprobs: null
  • model: "gpt-4"
• code: 200

Each example below simulates an unsuccessful API call to OpenAI, such as when the request has an unsupported format, or when a 404 or 500 error occurs. You can enable the "Path" option to handle failures within the flow.

Unsupported Format

This could occur if the API receives a non-JSON input or incorrect headers.

• outcome: failure
• outcomeReason: request-failed
• cause: "io.circe.ParsingFailure: expected JSON value, got '<!doctype...' (line 1, column 1)"
• message: "Outbound Webhook Failed: RawContentType[text/html] was not supported, and response could not be decoded as JSON\n. Response[code=200, content-type: Some(text/html), body=Right(<!doctype html><html><body>Error</body></html>)]"
• code: 200

404 Not Found

This example shows a 404 error, which might happen if the request tries to access a non-existent endpoint.

• outcome: failure
• outcomeReason: request-failed
• code: 404
• cause: "Failed outbound request: Not found"
• message: "Outbound Webhook Failed: Request failed with non-successful status code"

Internal Server Error

A 500 error example where the OpenAI API has a server-side issue.

• outcome: failure
• outcomeReason: request-failed
• code: 500
• cause: "Failed outbound request:"
• message: "Outbound Webhook Failed: Request failed with non-successful status code"

Use data from a web request response: Sample response

When using Make a web request, data is not only sent to an endpoint, but a response can also be received from the external service. The data from the response can be used in the flow by referencing merge fields in later nodes.

For a detailed guide on merge fields, see Working with merge fields

To view and use these merge fields, we need to provide the Make a web request node with a sample response.

By navigating to the 'Merge fields' tab on the Make a web request node, you will see the merge field explorer, 'This node' is selected, and the ability to paste a sample payload of the response:

The sample response must be in JSON format (key value pairs, objects, arrays). Here's an example of a response body, in JSON format:

{
"userid": "harrison-example",
"firstname": "harrison",
"lastname": "koenig",
"mobile": "+61400000000"
}

Note that any sample properties with 'null' values can't have a type, and sample value, inferred from them. Please only provide sample properties with not-null values.

When you paste the sample response into the field, Pendula checks to make sure the data is correctly formatted, and highlights any errors that may need to be corrected, so you can be confident the data is valid before starting the flow.

Note that we currently only support one sample payload. If your responses can have different structures, you may need to manually enter the merge fields by typing.

Merge fields

You can reference any merge fields from nodes previous to this node. Learn more about merge fields

In later nodes, when viewing the available merge fields for the Make a web request node in merge field explorer you can expect to see:

If your flow is using the Salesforce integration, the DATE value will be in ISO 8601 and DATE/TIME value will be in Unix Epoch. For more information about using a Salesforce integration node with the Format date node, refer to the document Working with Salesforce DATE and DATE/TIME fields

• Platform	Format Example	Format	System Type
  Salesforce	2025-03-28T14:30:00.000Z	ISO 8601	CRM & ERP
  SAP	2025-03-28T14:30:00.000Z	ISO 8601	CRM & ERP
  Oracle ERP & Oracle DB	2025-03-28 14:30:00	ISO 8601	CRM & ERP
  NetSuite (Oracle)	2025-03-28T14:30:00.000Z	ISO 8601	CRM & ERP
  Microsoft Dynamics 365	2025-03-28T14:30:00Z	ISO 8601	CRM & ERP
  Shopify	2025-03-28T14:30:00Z	ISO 8601	E-commerce & Payments
  Magento (Adobe Commerce)	2025-03-28 14:30:00	ISO 8601	E-commerce & Payments
  WooCommerce	2025-03-28T14:30:00	ISO 8601	E-commerce & Payments
  Stripe	1711649400 / 2025-03-28T14:30:00Z	Unix Epoch & ISO 8601	E-commerce & Payments
  PayPal	2025-03-28T14:30:00Z	ISO 8601	E-commerce & Payments
  Google Analytics (GA4 & UA)	2025-03-28T14:30:00Z	ISO 8601	Marketing & Analytics
  Facebook Ads (Meta Ads Manager)	2025-03-28T14:30:00Z	ISO 8601	Marketing & Analytics
  HubSpot	2025-03-28T14:30:00Z	ISO 8601	Marketing & Analytics
  Mailchimp	2025-03-28T14:30:00+00:00	ISO 8601	Marketing & Analytics
  Workday	2025-03-28T14:30:00Z	ISO 8601	HR & Payroll
  ADP (Payroll & HR)	2025-03-28T14:30:00Z	ISO 8601	HR & Payroll
  BambooHR	2025-03-28T14:30:00Z	ISO 8601	HR & Payroll
  Zendesk	2025-03-28T14:30:00Z	ISO 8601	Support & Communication
  ServiceNow	2025-03-28T14:30:00Z	ISO 8601	Support & Communication
  Jira (Atlassian)	2025-03-28T14:30:00.000+0000	ISO 8601	Support & Communication
  Slack API	1711649400	Unix Epoch	Support & Communication
  Microsoft Teams API	2025-03-28T14:30:00Z	ISO 8601	Support & Communication
  SAP Business One	2025-03-28T14:30:00Z	ISO 8601	Logistics & POS
  Square POS	2025-03-28T14:30:00Z	ISO 8601	Logistics & POS
  Clover POS	1711649400	Unix Epoch	Logistics & POS
  AWS S3 / CloudWatch	2025-03-28T14:30:00Z	ISO 8601	Cloud & Hosting
  Azure (Blob Storage, Functions, etc.)	2025-03-28T14:30:00.000Z	ISO 8601	Cloud & Hosting
  Google Cloud Storage & Logging	2025-03-28T14:30:00Z	RFC 3339 (ISO 8601-compliant)	Cloud & Hosting
  Google Calendar	2025-03-28T14:30:00Z	RFC 3339 (ISO 8601-compliant)	Calendars & Scheduling
  MySQL	2025-03-28 14:30:00	ISO 8601	Databases
  PostgreSQL	2025-03-28 14:30:00+00	ISO 8601	Databases
  SQL Server	2025-03-28 14:30:00.000 +00:00	ISO 8601	Databases
  BigQuery	2025-03-28 14:30:00 UTC	ISO 8601 & Unix Epoch	Databases
  Firestore	1711649400	Unix Epoch	Databases
  Google Sheets	2025-03-28T14:30:00Z (or serial number)	ISO 8601 & Google Serial Date	Databases

Selecting the output format

This section of the node enables you to select how you want a field to be formatted. You will have the option of selecting the type of date time you want to format and your desired output format of this data.

Output format
Choose the desired output format in this field. The dropdown menu provides commonly used date formats and the option to select a "Custom format.” Once selected, the "Input format token" field appears, allowing token formats to be entered from the table below. Separate multiple tokens with a comma or space to build out a desired format.

• Standalone token	Format token	Description	Example
  S		millisecond, no padding	54
  SSS		millisecond, padded to 3	054
  u		fractional seconds, functionally identical to SSS	054
  uu		fractional seconds, between 0 and 99, padded to 2	05
  uuu		fractional seconds, between 0 and 9	0
  s		second, no padding	4
  ss		second, padded to 2 padding	04
  m		minute, no padding	7
  mm		minute, padded to 2	07
  h		hour in 12-hour time, no padding	1
  hh		hour in 12-hour time, padded to 2	01
  H		hour in 24-hour time, no padding	9
  HH		hour in 24-hour time, padded to 2	13
  Z		narrow offset	+5
  ZZ		short offset	+05:00
  ZZZ		techie offset	+0500
  ZZZZ		abbreviated named offset	EST
  ZZZZZ		unabbreviated named offset	Eastern Standard Time
  z		IANA zone	America/New_York
  a		meridiem	AM
  d		day of the month, no padding	6
  dd		day of the month, padded to 2	06
  c	E	day of the week, as number from 1-7 (Monday is 1, Sunday is 7)	3
  ccc	EEE	day of the week, as an abbreviate localized string	Wed
  cccc	EEEE	day of the week, as an unabbreviated localized string	Wednesday
  ccccc	EEEEE	day of the week, as a single localized letter	W
  L	M	month as an unpadded number	8
  LL	MM	month as a padded number	08
  LLL	MMM	month as an abbreviated localized string	Aug
  LLLL	MMMM	month as an unabbreviated localized string	August
  LLLLL	MMMMM	month as a single localized letter	A
  y		year, unpadded	2014
  yy		two-digit year	14
  yyyy		four- to six- digit year, pads to 4	2014
  G		abbreviated localized era	AD
  GG		unabbreviated localized era	Anno Domini
  GGGGG		one-letter localized era	A
  kk		ISO week year, unpadded	14
  kkkk		ISO week year, padded to 4	2014
  W		ISO week number, unpadded	32
  WW		ISO week number, padded to 2	32
  ii		Local week year, unpadded	14
  iiii		Local week year, padded to 4	2014
  n		Local week number, unpadded	32
  nn		Local week number, padded to 2	32
  o		ordinal (day of year), unpadded	218
  ooo		ordinal (day of year), padded to 3	218
  q		quarter, no padding	3
  qq		quarter, padded to 2	03
  D		localized numeric date	9/4/2017
  DD		localized date with abbreviated month	Aug 6, 2014
  DDD		localized date with full month	August 6, 2014
  DDDD		localized date with full month and weekday	Wednesday, August 6, 2014
  t		localized time	9:07 AM
  tt		localized time with seconds	1:07:04 PM
  ttt		localized time with seconds and abbreviated offset	1:07:04 PM EDT
  tttt		localized time with seconds and full offset	1:07:04 PM Eastern Daylight Time
  T		localized 24-hour time	13:07
  TT		localized 24-hour time with seconds	13:07:04
  TTT		localized 24-hour time with seconds and abbreviated offset	13:07:04 EDT
  TTTT		localized 24-hour time with seconds and full offset	13:07:04 Eastern Daylight Time
  f		short localized date and time	8/6/2014, 1:07 PM
  ff		less short localized date and time	Aug 6, 2014, 1:07 PM
  fff		verbose localized date and time	August 6, 2014, 1:07 PM EDT
  ffff		extra verbose localized date and time	Wednesday, August 6, 2014, 1:07 PM Eastern Daylight Time
  F		short localized date and time with seconds	8/6/2014, 1:07:04 PM
  FF		less short localized date and time with seconds	Aug 6, 2014, 1:07:04 PM
  FFF		verbose localized date and time with seconds	August 6, 2014, 1:07:04 PM EDT
  FFFF		extra verbose localized date and time with seconds	Wednesday, August 6, 2014, 1:07:04 PM Eastern Daylight Time
  X		unix timestamp in seconds	1407287224
  x		unix timestamp in milliseconds	1407287224054

Example custom output formats

Time zone
Select the timezone for the output date time. The chosen UTC offset adjusts the date and time accordingly. For example, selecting UTC+11:00 converts the output to Australian Eastern Daylight Time (AEDT) during daylight saving or UTC+10:00 for Australian Eastern Standard Time (AEST) outside of daylight saving.

The Upload from Base64 action is an Early Access feature. Contact your Account Executive to have it enabled for your Pendula tenant.

What is Upload from Base64?

The Upload from Base64 action processes a Base64 array and makes it available as a formatted usable file to reference in your flow. This is achieved by providing the base64 code, file name (including the extension type), and the mimeType. The mimeType is what informs the action of the type of file it should create.

Node configuration

Sending base64 files to the flow

Webhook trigger: array schema

The input for this array will need to be included as a payload in the Merge fields tab of the Webhook trigger at the start of the flow. 

When structuring your trigger payload, you must include the file array of Base64 code, the name of the file (including the extension), and mimeType. 

You can nest multiple files in the payload array under a single merge field that can be referenced throughout the flow. For example, when this merge field is referenced in the email attachment node, it will send both attachments. In the following example, we are creating a calendar invite attachment and a PDF for an email:

file: The content of the file encoded in Base64
name: the name of the file including the extension
mimeType: the mime type of the file.

Example JSON payload

{
  "attachments": [
    {
      "file": "base64 code goes here",
      "name": "file name including extension.ics",
      "mimeType": "text/calendar"
    },
    {
      "file": "base64 code goes here",
      "name": "file name including extension.pdf",
      "mimeType": "application/pdf"
    }
  ]
}

• Extension	File Type Description	MIME Type
  .txt	Plain text	text/plain
  .html	HTML document	text/html
  .jpg/.jpeg	JPEG image	image/jpeg
  .png	PNG image	image/png
  .gif	GIF image	image/gif
  .webm	WebM video	video/webm
  .mp4	MPEG-4 video	video/mp4
  .mpeg	MPEG video	video/mpeg
  .pdf	PDF	application/pdf
  .doc	Microsoft Word document	application/msword
  .docx	Word Open XML document	application/vnd.openxmlformats-officedocument.wordprocessingml.document
  .xls	Microsoft Excel spreadsheet	application/vnd.ms-excel
  .xlsx	Excel Open XML spreadsheet	application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

For details of supported file types and other useful information about working with attachments in Pendula please see: Outbound email node (with attachments)

Settings

In the Upload from Base64 node, the settings tab contains a single input field. The input field requires an array matching the schema above to be supplied.

1. In the settings tab of the node, select the "Input Base64 array" field. 

2. In the drop-down list, select the "attachments" merge field. The merge field name is the same as the object array name from the Webhook trigger payload. This contains the file, mimeType, and file name.

   

3. Click "Save" and exit the node.

Outcomes

Outcome paths can be configured if you wish for the action to perform a different next step based on whether it was successful or not. The following outcomes are available for this node:

Success: the action was successful in uploading the file/s

Success: No Files: an empty array was supplied and no files were included. This allows a single flow to handle different use cases where some recipients may receive attachments while others don't.

Failure: the action was unable to upload the files. This may be due to a misconfiguration or the supplied data not meeting the requirements.

Use case: sending the attachments as emails in the flow

The file/s are now available to include in a subsequent Outbound Email node. 

In the Attachments tab of the Outbound Email node, use the Search attachment fields lookup to select the attachments from the previous node.