Open the scenario editor by going to Campaigns > Scenarios > Create new.

Adding, editing, and connecting components

Every scenario is a set of components connected to a flow.

• To add a component, click on the component you want to add, or you can drag & drop it onto the board with your left mouse button.
• To edit a component, double-click the component with your left mouse button. This will open a component-specific edit window. Click "save" after you are done.
• To remove a component, select the component by clicking it with your left mouse button and then click on the trash icon, or hit backspace or delete on your keyboard.
• To connect two components, click on the output connector on the right side of the first component and drag the line to the input connector on the left side of the second component, as shown in the animation below. Release your mouse once connected.
  • Now you have created a flow. Customers will pass through this flow starting from a trigger (e.g.,On date) to operators (e.g.,Condition) and finally into actions (e.g.,Email).

📘

While in every meaningful scenario there should be at least one trigger and one action component, there can be multiple flows in a single scenario.

🚧

Scenario size recomendations:

There is a hard limit for how much data can be stored in one scenario (or email campaign) definition.
Following statements are taking into account average size of scenario node.
Using more nodes can quickly lead to internal error about scenario size limit.

• Scenario should not use more than 100 nodes. Ideally, keep their number below 30 to avoid possible lags.
• Scenario should not use more than 25 email nodes. This includes different language versions. The shorter your email is, the more nodes or translations you can use. As a workaround you can move parts of your email into html blocks. Content of such reference will not count towards given scenario size limit.
• Maximum scenario size is 7.9 MB.

Copy & paste nodes

The nodes and their connections can easily be copied within one scenario or among multiple scenarios.

This can be done by selecting the nodes you want to copy, clicking Cmd/Ctrl + c, and then clicking Cmd/Ctrl + v to paste (hold Shift if you want to select multiple nodes).

When copying from one scenario to another for the first time, the browser might ask to allow seeing the nodes copied to the clipboard.

🚧

Copy & pasting between scenarios is not allowed between different versions of the app.

Moving, zooming, and selecting multiple components

You can move your position on the canvas by clicking on an empty space and dragging your mouse. To select multiple components, hold shift and do the same. You can delete all selected components by pressing backspace on your keyboard. Zoom in or out using your mouse wheel.

Using a touchpad, you can move by sliding 2 fingers and zoom in/out by pinching 2 fingers.

Keyboard shortcuts available:

• Press 1 to fit the screen
• Press 2 to zoom into the selection
• Press 0 to reset zoom.

You can also click on the minimap in the bottom right corner to navigate your position.

🚧

Saving and executing a scenario

Don't forget to save your scenario by clicking the "Save" button in the top right corner. This will save the scenario for later and will not launch it. If you want to launch the scenario now, click the "Start" button. This means that the triggers in the scenario will now start streaming customers into the flow.

Scenario components explained

Let's have a look now at how each component works. Remember that you can specify each component by double-clicking on it once it has been placed on the canvas.

How are components displayed

Names and search

The name of the node displayed in the Scenario contains the following 3 things:

• The Action ID of the node
• The type of the node
• The user-defined name of the node

Search is available to help clients quickly find the node they are looking for. It is possible to search for the node by its action ID, the type of the node, or its name.

Labels formatting

Labels can be formatted using standard markdown notation. This can be done using either the visual formatting helper bar or by writing markdown directly. The available formatting options are:

• bold
• italic text
• headings (H1-H6)
• unordered lists
• links

Other markdown formatting options are also supported but might not be properly styled in the resulting label.

Triggers

Trigger limitations

Certain campaign event statuses do not trigger scenarios when used in the "On event" node. Therefore, you should not use them as triggers in your scenarios. Those statuses include:

• enqueued
• enqueue_failed
• sent
• failed

Importantly, campaign events that do not have a status attribute altogether, such as A/B tests, will also not trigger a scenario.

Apart from campaign events, there are other events that do not trigger scenarios. Those are listed below:

• merge
• consent event that is generated by manual consent update in the web UI

Additionally, unless the option "Allow scenario triggering" in the add event node is turned on, events added/generated from scenarios don't trigger scenarios with triggers waiting for that particular event.

Resetting triggers

Once used in a live scenario, both Now and On date triggers will become inactive. If you want to re-activate them, use the Reset trigger. In the picture below, you can see the triggers as active on the left and inactive on the right, with two highlighted options on how to reset them:

🚧

Specifying events using derived attributes

It is not possible to specify events using derived attributes (running aggregates, expressions, event segmentations).

❗️

Parallel triggers overload in a single scenario

As the triggered actions will be processed in the order as requested, a single Scenario with parallel triggers might not work as expected. For example, if both 'On event' and 'On date' triggers are pulled simultaneously the 'On event' trigger would need to wait until the 'On date' trigger is fully processed despite the expectation that it would go into action right after a particular event is tracked. If there are multiple parallel triggers it is recommended to divide them across multiple Scenarios.

Actions

When a customer enters an action component, an action is executed. Each type of action has a unique 'action_id'. Read more about this in the System events article.

📘

Locating action ID

To see a particular action ID, go to the Evaluation tab, hover your mouse over the node and the ID will be shown in the format of "#(number)".

🚧

Emailing a manually-chosen recipient

If the recipient is used in the scenario's email node is chosen manually, then the campaign policy for the frequency of delivery will be ignored.

Operators

Operators are generally used to change the way customers move through the flow. There are the following types of Operators:

📘

Relative time filters use the UTC timezone while absolute time filters use your custom timezone from Settings > User Settings > Timezone. Note that the relative filters apply only to scenarios, not to your whole project.

Flexible operators

"Set attribute," "Add event," and "Condition nodes" can be defined not only by using a visual editor/customer filter but also by using their code editor. JSON syntax has to be followed, and jinja can be used for further personalization and to define a condition.

The following are some syntax examples that can be used in the code editor:

Add event node:

[
	{
		"event_type": "event", 
		"timestamp": 1234,
		"properties": {
			"property1": "value1",
			"property2": 1234,
			"property3": true,
			"property4": ["value1", "value2"],
			"property5": {"key1": "value1"},
			"property6": {{ variable | json }}
				// json filter should be set up to keep the Jinja personalization valid for the app
				// (e.g., to ensure that string values are surrounded by double-quotes instead of single-quotes
				// or boolean values are true/false instead of True/False)
		},
		"trigger_scenarios": false
	}
]

Set attribute:

{
	"property1": "value1",
	"property2": 1234,
	"property3": true,
	"property4": ["value1", "value2"],
	"property5": {"key1": "value1"},
	"property6": {{ variable | json }}
		// json filter should be set up to keep the Jinja personalization valid for the app
		// (e.g., to ensure that string values are surrounded by double-quotes instead of single-quotes
		// or boolean values are true/false instead of True/False)
}

Condition:

{% if customer.first_name %}
	True
{% else %}
	False
{% endif %}

or

// Following Jinja will be evaluated as True
-------BEFORE RENDER--------
{% set x = True %}
{{ x }}
--------RENDERED------------
"True"
----------------------------

// Following Jinja will be evaluated as False
--------BEFORE RENDER-------
{% set y = False %}
{{ y }}
--------RENDERED------------
"False"
--------BEFORE RENDER-------
{% set z = 123 %}
{{ z }}
--------RENDERED------------
"123"
----------------------------

🚧

It is not possible to test preview customers flowing through the Jinja Condition node to the following nodes and therefore, N/A will be displayed in the Test tab instead.

Customer limit

🚧

You can only have up to 1000 distinct groups in a limit node, before it starts failing.

The customer limit allows you to set the maximum number of customers who can pass through the node. This can be very useful when your scenario leads your customers to a channel with a limited capacity. For example, you might have limited capacity for the number of calls your call center can handle within a day.

The customer limit node creates two outputs: a successful one for the customers within the limit and the unsuccessful one for the customers over the limit. It is at the moment when a particular customer arrives at the customer limit node in the scenario that it is decided to which output they will go. At first, all customers go to the successful output, but once the threshold is reached, all the rest go to the unsuccessful output. Since then, all customers will be falling into the unsuccessful output even for the next campaign runs until the number of customers passing through the node is reset back to zero. You can set a period reset of this number in the Specify customer limit reset period.

You can set the limit by filling in the value of the Limit to # of customers. This value can either be an absolute number or a number referenced by jinja (references to the following can be used: metrics, reports, report_value_by_key, catalogs). Read more on jinja references.

You can also Allow drill-down group which lets you use jinja to add multiple limits per node based on the group - a separate limit on each segment of customers using the scenario. You might want to do this if you send out a competition that would be decided on a 'first come, first serve' basis, but where you also want to have a specific number of winners from different segments of customers. You can use the same variables as in webhooks, like customer attributes, trigger event properties, or webhook response. If you want to use the drill-down group as a parameter in the jinja of the limit, the parameter is called limit_group.

📘

Example of the limit with a limit group

To look into catalog subsidiaries and look for an item that was used as a drill-down group, and reference limit attribute from the catalog item, use:
{{ catalogs.subsidiaries.item_by_id(limit_group).limit }}

Using optimal send time

Based on your customer's past behavior, the optimal send time option in the wait node allows you to automatically send out email campaigns to individual customers at a time, rounded to an hour, when they are most likely to click/open your email. Learn more about Optimal send time

Using Dynamic Time

You can use the dynamic wait time to set a custom wait period using jinja. You can use some project metrics, aggregates, trigger event attributes, or any type of personalization.

Common usage of the dynamic wait time is to spread or throttle the campaign execution over some time period using the random number generator. E.g., to spread the send approximately within 60 minutes, you can use the{{ range(0,60) | random }}.

If the Wait time is set to hours, the number is rounded down (to an integer). For example, when you calculate 1.8 hours wait, Wait node waits only 1 hour. If you want to round up, you can use the Jinja function round (can accept "ceil", "floor" or "common" parameter):

Wait {{ ((event.departure_time - time) / 3600) | round("ceil") }} hours

Wait node does not wait at all if the wait value is zero or lower (executes the following node right away).

Delivery of an SMS campaign can be either successful or unsuccessful. Bloomreach Engagement tracks this information as an attribute of the campaign events.

Using Silent Hours

You can use Silent hours to temporarily pause the sending of your campaigns to the customer during times when you do not want to disturb them, such as at night or on a Sunday. If this node is entered during silent hours, the scenario will wait under their end and only then send the campaign. Silent hours can be set up for a specific time repeated daily or as Silent days, which will silence the whole selected day, or both options can be used together. A specific timezone can also be selected for the silent hours to be evaluated against.

Silent hours can be also set up globally in Project settings > Campaigns > General Campaigns settings. This will pre-fill a Silent hour Wait node when it would be selected in a scenario. The setting can be overridden in a specific node if needed.