Marketing Cloud Intelligence Custom Classification are attributes of the “Main Entity Key” of each data stream type. Note, one Main Entity key (belonging to any data model) can hold only one Custom Classification value.
For example, let’s say you own a sporting goods brand, and these are the three channels that you leverage in your overall marketing strategy:
1. Paid Search
2. Social Media
3. Email Marketing
All the data sources belonging to these channels are ingested, harmonized, and visualized in Marketing Cloud Intelligence. Thanks to Datorama’s data model, various API connectors, and multiple widget types (interactive, element, and custom) it allows for the building of award-winning visuals.
Now, let’s add a little complexity here. Let’s assume that there’s a requirement to build a multi-brand dashboard that can slice your Paid Search, Social and Email marketing data by Brand and Activity. Since you already have a fully functional Marketing Cloud Intelligence workspace, with APIs scheduled to pull data and dashboards for each channel, you would naturally prefer supplementing your primary data sources with additional information like Brand or Activity, without making drastic changes to the back-end setup.
Here’s where the Marketing Cloud Intelligence Custom Classification entities shine. Since these entities are available across all data models, you can easily pivot your Search, Social, and Email data using these entities. Now, how do you map these entities in your existing data streams?
Before you proceed with mapping these entities, these are a few things to check:
If all your data sources meet the above criteria, then you can proceed with mapping Custom Classification entities in each of your data streams. MCI offers five Custom Classifications levels in every data model, as shown in the image below:
Each Custom Classification level has 20 attributes attached to it. These attributes are labelled as Levels, so there are 20 Levels in each Custom Classification, as seen in the image below:
The relation between Main Entity Keys (of all Data Models) and Custom Classification 1 to 5:
A Main Entity Key has one-to-one relation with Custom Classification 1 to 5. It means one Main Entity Key (Let’s say Media Buy Key from Ads data Model) can hold only 1 value recorded in Custom Classification 1 to 5.
There’s a Many to Many relation between Custom Classification 1, Custom Classification 2, Custom Classification 3, Custom Classification 4 and Custom Classification 5.
Now that we understand how Custom Classification works in MCI, let’s plan the changes needed in Social, Search, and Email marketing data streams to supplement them with Brand and Activity level information.
Firstly, we will modify our existing Google Ads, Facebook Page, and Email marketing data streams to map in Custom Classification 1 and 2. Custom Classification 1 will store the Brand Code and Custom Classification 2 will store the Activity Code. Once mapped, we will re-run the streams.
We will create two additional data streams with only the Custom Classification fields mapped in it. These two streams can belong to any data model since Custom Classification fields are available across all data stream types. One of the data streams will solely handle Brand Classification with “Brand Code” & “Brand Name” and the other will handle the Activity classification with “Activity Code,” “Activity Name,” and “Activity Description.”
This flow diagram gives a bird’s eye view of how Custom Classification fields will help slice and dice data that belongs to different channels.
Brand Classification table preview:
Activity Classification table preview:
Once the data streams with Custom Classification fields are processed, you’re all set to pivot your Search, Email, and Social data with Brand names, Activity Name, and Description.
Preview of Search, Social, and Email Marketing metrics pivoted with Activity & Brand information:
Of course, the same results can be achieved using other harmonization features in MCI’s Harmonization Center, including Data Fusion, Parent Child, Empty Parent, Patterns, and Data Classification. But, Custom Classification entities are comparatively easier to manage and saves a lot of time if you need to supplement primary data sources with additional information without making large scales changes in your back-end setup.
Custom Classification entities record data at a workspace level, so deleting the data stream doesn’t delete the Custom Classification’s dimensional values. Since the dimensions recorded with Custom Classifications aren’t tied to a measurement, they don’t consume any rows in your workspace. If you need to delete Custom Classification’s dimensional values, you must manually remove them via the Dimension Explorer in your workspace.
About Decision Foundry
Decision Foundry is a Salesforce, independent software vendor, managed services provider, and a certified award-winning Salesforce Marketing Cloud integration partner. Decision Foundry closes the gap between data accessibility, platform adoption and business impact. Our consulting services include the integration of Data Cloud, Account, Engagement, Personalization, Tableau, and Intelligence.