Introducing Facebook as an Action: Reduce customer acquisition cost & increase retention by using the output of your Machine Learning models as Custom Audiences on Facebook Ads


A great way to reduce your advertising cost on Facebook is to create a Custom Audience. To do so you would most likely start targeting ads to people who’ve already visited your site. That’s great if you want to act based on data only Facebook collects. But what about other sets of data, such as information extracted by a CRM like Hubspot, or a cloud warehouse like Snowflake, and most importantly, the outcomes from Machine Learning models built from those data sources? Datagran does just that.


As part of our Pipeline tool, today we introduced Facebook as an Action. With this brand new element, you can now export custom audiences from ML models ran inside Datagran directly to Facebook ads, to reduce CAC, increase revenue, and more. 


“The whole idea of custom audiences is to target your audiences better. The result is to save money and to be more efficient in your ad spend. For example, when we use a custom audience on Facebook, our cost goes down by 72%”. Carlos Mendez, Datagran’s Co-Founder & CEO.


To start creating custom audiences:


1. Sign in to your Datagran account and head over to Pipelines.

2. Start a Pipeline, by adding the data sources you want to extract information from, for example: MySQL, Shopify, Salesforce.

A data workflow dashboard.

3. Drag and drop a Custom SQL Operator if you want to see what’s inside of your data source. With this element, you can add or remove any column. If you don’t know coding, the Editor helps you clean up your data without the need to write any code.


4. Once that is done, add Export Audience to Facebook Action into the canvas. Hit the Edit button.


Before this action is run, make sure your pipeline is providing as many identifiers as possible. These are used to match users, and the more specific, the better the match. Some examples of identifiers are: Name, email, gender. Run your Pipeline and save it. 


6. Locate your Facebook business account. 


7. Choose your Facebook Ad Account from the drop down menu.


8. Name the target audience.



9. Select the dataset you want to extract information from.


10. Create a Custom Audience by adding a customer value column. Adding a customer value column allows you to create value-based lookalike audiences later on.



11. Preview and map out your data from map column to identifier to ensure formatting is correct. Click on each tab to view each one.



11. Now that you have specified your data identifiers, click Save and Run your Pipeline. 

12. Your Custom Audience will be pushed to your Facebook Ads account, and it can be viewed inside of your Facebook Business Manager Audience List. The processing time for the audience to be exported to your Facebook Account may vary depending on the volume of your dataset.

Check out our guide to build Pipelines for some quick examples with Playground invoice and customer data, Custom SQL, and more.