Send propensity-based customers to Facebook

Published August 16, 2022

Modules

Difficulty

Industry

With Synerise you can choose very specific segments of customers for your marketing communication on Facebook and spend less on your campaigns.

In this use case, you will use the Predictions module to calculate the probability of customers buying items from shoes category. The results will be later used in the conditions of segmentation that will trigger the workflow. In the workflow we will use the native Facebook integration to send customers with highest propensity to buy, as a Custom Audience. This group of customers can be later used for better targeting of your marketing efforts.

Prerequisites

Enable the Propensity prediction type.
Generate the system user access token in Facebook; as the scope value, select ads_management.
Create Custom Audience in Facebook
Implement transaction events using SDK or API.
Create an item feed.
The category attribute must be added to filterable attributes.

Process

In this use case, you will go through the following steps:

Create a Propensity prediction that produces the 2-point score (low and high).
Create a segmentation of customers.
Create a workflow.

Create a Propensity prediction

In this part of the process, you will create a propensity prediction to purchase any product from the shoes category for the audience of recognized customers (those that have the email attribute). In further steps, an email address will be used as a customer identifier while sending audience to Facebook.

Select the model type

Go to > New prediction.
In the Select prediction type window that appears, click Propensity.
Cllick Apply.

Select customers to be analyzed

Select the audience for whom you want to prepare a prediction.

In the Audience section, click Define.
Click Choose segmentation.
On the dropdown list, click Create new.
In the Segmentation name field, enter a meaningful name of the segmentation.
Click Next step.
Click Choose filter:
1. From the dropdown list, select the email attribute.
2. As the logical operator, select Regular expression.
3. Type ..
4. Save the segmentation by clicking Create segmentation.
Click Apply.

The view of propensity audience configuration — Propensity audience configuration

Define the item

In this section, you define the category for which you want to calculate the prediction, in our case it’s the shoes category. This is done by creating a filter that matches the category in the catalog.

In the Item selection section, click Define.
Click Choose item feed.
Select the catalog that contains the items you want to make the prediction for.
Result: The Item filter section appears.
Click Define item filter.
From the Select attribute dropdown list, select the category attribute.
As the logical operator, select Equal.
Click Select category and select the desired product category - men> shoes & accesories> all shoes You can use the search field.
In the Level range field, enter 0. Enter 0 for all shoes.
Click Add filter. Choose the item feed. From the Select attribute drop-down list, select the category attribute.
As the logical operator, select Equal.
Click Select category and select the desired product category - women> shoes & accesories> all shoes You can use the search field.
In the Level range field, enter 0. Enter 0 for all shoes.
Change the AND operator to OR.
Click Save.
Save the item feed configuration by clicking Apply.

The view of propensity item feed configuration — Propensity item feed configuration

Additional settings and saving

By default, the calculation is performed once and the results are presented on a 5-point scale.

If you want to leave the settings as default, the predictions is ready to calculate. Continue to saving the prediction.
If you want to change the settings, in the Settings section, click Change.

To change the scale from 5-point to 2-point, select the 2-point scale radio button.
Click Apply.

Result: After the calculation is complete a snr.propensity.score event is saved in the profiles of each customer in the audience. The event data includes detailed results of the prediction. Based on the snr.propensity.score event, you can create segmentations of customers with different propensity.

Create a segmentation

Based on the snr.propensity.score event, create a segmentation of customers with high propensity to purchase any item from the shoes category.

Go to Analytics > Segmentations > New segmentation.
Enter the name of the segmentation.
Click Performed action…
Select the snr.propensity.score event.
Click + where.
As the event parameter, select modelId (available in the parameters of the snr.propensity.score event).

The wiev of properties of the snr.propenisty.score event — Properties of the snr.propenisty.score event

As the logical operator, select Equal.
In the text field, enter the value of the modelId parameter.
Click + and where.
As the event parameter, select score_label.
As the logical operator, select Equal.
In the text field, enter high.
Set the date range according to your buisness needs.
Click Save.

The view of segmentation configuration — Segmentation configuration

Create a workflow

As the final part of the process, create a workflow that sends customers with the high propensity to buy items from the shoes category to Facebook.

Go to Automation > Workflows > New workflow.
Enter the name of the workflow.

Choose the segmentation of customers

As the first node of the workflow, add Audience. In the node settings:
1. In the Define audience section, select the segmentation you created in the previous step, click Apply to confirm.
2. Confirm by clicking Apply.

Configure the Facebook integration

On the Audience node, click then.
From the dropdown list, select the Add profiles to Custom Audience node.
Click the node.
From the Select connection dropdown list, select an existing connection which allows you to authorize in Facebook Ad.
If you haven’t established a connection yet:
1. At the bottom of the dropdown list, click Add connection.
2. In the Access token field, paste the token generated as a part of prerequisites.
3. Click Next.
4. In the Connection name, enter the name of the connection (it will be only visible on the list of connections).
  Tip: The connection can be re-used to any custom audience created in the Facebook Ad Account related to the token generated as a part of prerequisites.
In the API version field, enter the Facebook API version according to the Facebook documentation.
From the Identified type dropdown list, select Email.
In the Identifier source field, enter {{ client.email }}.

Tip: You can read more about Jinjava tags in the Automation module here.
In the Audience ID field, enter the ID of the Custom Audience to which you want to send your group of customers.
Confirm by clicking Apply.

Add the End node

On the Add profiles to Custom Audience node, click then.
From the dropdown list, select End.
In the upper right corner, click Save & Run.

Important: Wait a few minutes for the response of the webhook.

Result: An automation event is saved to the profiles of the customers. Then, go Facebook, to Facebook Manager > Audiences to check your Custom Audience.

The view of workflow configuration — Workflow configuration

What’s next?

You can start creating personalized ads to your customers in Facebook. You can see your custom audience in Facebook Manager > Audiences.

Check the use case set up on the Synerise Demo workspace

You can check the configuration of every element of this process directly in Synerise Demo workspace:

If you don’t have access to the Synerise Demo workspace, please leave your contact details in this form, and our representative will contact you shortly.