Create Data Views

19 min

Objectives

Understand the Data View UI
Understand the basic settings of session definition
Understand Attribution and Persistence within a Data View

Data View

With your connection done, you can now progress to influencing visualization. A difference between Adobe Analytics and CJA, is that CJA needs a Data View in order to clean and prepare the data before visualization.

A Data View is similar to the concept of Virtual Report Suites in Adobe Analytics, where you define context-aware visit definitions, filtering and also, how the components are called.

You’ll need a minimum of one Data View per Connection. However, for some use-cases, it’s great to have multiple Data Views for the same connection, with the goal of giving different insights to different teams. If you want your company to become data-driven, you should adapt how data is viewed in each team. Some examples:

UX metrics only for the UX Design team.
Use the same names for KPIs and Metrics for Google Analytics as for Customer Journey Analytics so that the digital analytics team can speak one language only.
Data View filtered to show, for instance, data for one market only, or one brand, or only for Mobile Devices.

To create a new dataview:

Navigate to Data Management > Data views .
Click on the icon Create new data view.

You'll be redirected to the Create Data View workflow.

Data View Definition

You can now configure the basic settings for your Data View.

Select the Connection you created in the previous exercise from the Connection drop down.

If you did not create a new Connection, you can use the Connection created by us: Currys - AEP Bootcamp Data Connection .

Give your Data View a name following this naming convention: [Your name] - Cross-channel Data view
For the Calendar, select the timezone Greenwich Mean Time; Monrovia, Casablanca [GMT]. Leave the rest as it is. This is a really interesting setting as some companies operate in different countries and geographies. Allocating the right time zone for each country will avoid typical data mistakes such as believing that for instance, in Peru, the majority of the people buy T-shirts at 4:00 AM.
You can also modify the main metrics naming (Person, Session and Event). This is not required, but some customers like to use People, Visit and Hit instead of Person, Session and Event (default naming convention from Customer Journey Analytics).

You should now have the following settings configured:

Select Save and Continue, then click on the Components tab.

Data View Components

In this exercise, you'll configure the components you need to analyse the data and visualise it using Analysis Workspace. In this UI, there are three main areas:

left side: available components from the selected datasets;
middle: added components to the Data View;
right side: component settings.

You now have to drag and drop the components you need for the analysis to the Included Components. To do this, you need to select the components in the left menu and drag and drop them onto the canvas in the middle.

Let’s start with the first component: Product Name (productListItems.name):

in the "Schema fields" box click on Event datasets > productListItems
drag and drop the field Name onto the canvas in the Dimension section.

Now you can configure the component settings:

click on the component Name, recently added to the dimension list; the section on the right will get populated with the component settings;
in the Component name text field type Product Name, since this component stores the name of the product purchased

Tips for Adobe Analytics users

something really important is the Persistence settings. The concept of eVars and prop doesn’t exist in CJA but the Persistence settings make a similar behaviour possible.

If you don’t change these settings, CJA will interpret the dimension as a Prop (hit level). Also, we can change the Persistence to make the dimension an eVar (persist the value across the journey).

Let’s leave the Page Name as a Prop. As such, you don’t need to change any Persistence Settings.

Attribution settings on metrics can also be changed in Analysis Workspace. In some cases you may choose to set it here to avoid business users from having to choose which is the best persistence model.

Next, you’ll have to configure all of your Dimensions and Metrics that you want in your Data view. Since we have limited time in our Hands-on Lab we will add all Dimensions and Metrics to our DataView.

Ensure Contains data is applied as a filter under the search bar.
Ensure that search text field is empty.
Select Add all underneath Schema Fields.

Your configuration should then look like this (Note, you may want to make the names of some variables more user friendly):

Component Settings

In some cases you may want to create new components based on the values captured in your variables. For example, when someone places a call with the call centre we could capture an EventType value of in our Experience Event data being ingested.

We want to create a component that will provide a count of call centre dial-ins in our Workspace. To create this:

search for Event Type;
drag the Event Type variable onto the Components canvas, into the Metrics panel;
- if you will see a 'Duplicate field' warning, select Add anyway. The Event Type component will be added at the top of your metrics;
click on this new component, we will alter the settings in the right hand panel;
add the component name: Event: Call center dial-ins ;
scroll down and expand Include Exclude Values;
check Set include/exclude values;
For the criteria, specify Equals callcenter.dialIns

This condition will ensure this variable when used in a Workspace, will only include only a count of events that have an EventType value of callcenter.dialIns.

We can repeat this for all of our possible events; ad clicks, email deliveries, call centre dial-ins, offers viewed, etc.

Let’s create another metric that we’ll reuse in an exercise later.

In Search schema fields, search for productView.
Drag and drop the item commerce.productViews.value in the Metrics list
Rename the metrics as Product Views.

To avoid loosing the work done so far, click Save.

Create Derived Fields

This feature helps you improve the quality and accuracy of data in your projects due to data errors or gaps.

Example use cases include:

Configure a derived page name field that corrects page name values that were improperly collected.
Configure a derived marketing channel field that determines the proper marketing channel based on one or more conditions, such as URL parameter, page URL, or page name.

Derived fields can save a significant amount of time and effort, compared to transforming or manipulating your data in other locations outside of Customer Journey Analytics. Such as Data Prep, Data Distiller, or within your own Extract Transform Load (ETL) / Extract Load Transform (ELT) processes.

As an example, we will create a derived field to give friendly names to our Dataset IDs. Using the standard Platform Dataset ID, derived fields allow for any custom string value friendly name identifier to be additively set as a component dimension.

From your Data view, select Create derived field from the left rail;
to define your derived field, select the Case When function from the functions list, and drag this onto the canvas.
In the Rule 1 configure the following settings: 🟡 add a screenshot with the proper datasetID and put the datasetID also here to be copied and paste 🟡

you can continue to add rule conditions for all of your Datasets into this derived field. Do this by selecting Add Condition on the canvas;
name your Field in the Field settings in the right hand rail as Dataset Name;
click Save.

You have now create a new derived field, you can now drag this into your Component Dimensions list.

Once you have added all the necessary components to your data view, click Save and continue.

Create a Derived Field for Bounces

We want to create a function template that uses a collection of rules to identify site bounces.

From your Data view, select Create derived field from the left rail;
to define your derived field, select the Case When function from the functions list, and drag this onto the canvas.
In the Rule 1 panel configure the following settings:
- If
  - Value: Session Starts, Criterion: Is set
  - Select the + icon to add a condition to this rule
  - In the added AND condition configure:
    Value: Session Ends, Criterion: Is Set
- THEN set value to
  - Value: Custom Numeric Value, Custom Numeric Value: 1
In the Otherwise set value to panel configure the following settings:
- Value: Custom Numeric Value, Custom Numeric Value: 0
This will mark the current hit as a bounce if a session starts and ends on the same hit.

Your final rule should look like:

Data View Settings

Click on Save and Continue and you should be redirected to the Data view Settings screen.

In this tab, you can modify some important settings to change how data is processed. Let’s start by setting the Session Timeout to 30 min. Thanks to every experience event’s timestamp you can extend the concept of a session across all channels. For instance, what happens if a customer calls the call center after visiting the website? Using custom Session Timeouts you have lots of flexibility in deciding what a session is, and how that session will merge data together.

In this tab you can modify other things like filtering the data by using a segments (you won’t need to do that in this exercise)

In this case, we will keep the default settings;
once you are done, click Save and finish.

Congratulations! You've completed the exerices and created the Data view. Now you've everything you need to analyse your data.

You can come back to this Data View afterwards and change settings and components at any time. Changes will affect how historical data is shown.