In the Report Groups step, you can configure whether the dimensions of selected report groups are recorded with the event and its captured values. Report groups enable multi-dimensional analysis of the attributes of an event.
A report group is a collection of dimensions, each of which may contain a different set of context data that is detected when the event is triggered. These attributes are recorded along with the value of the event. For example, the condition causing an event to fire could be that a user went to the checkout page. The value of the event could be the amount of the shopping cart. The attributes could be browser type, OS type, credit card type, and others.
- These individual attributes are called dimensions.
- Groups of dimensions are called report groups.
In the Report Groups step, the panel of a report group may be highlighted in gray. These report groups were disabled due to a high number of instances of the dimension (called facts) recorded for this report group over a one-hour interval.
Tealeaf enforces a global limit of 500,000 facts that can be recorded for a specific event over a one-hour interval. If this limit is exceeded, the fact is disabled, as well as any scorecards or dashboards that use the fact. This limit is imposed to prevent runaway fact creation and needless growth of the database.
A disabled fact must be re-enabled through the event definition.
Note: Before a fact is re-enabled, you should resolve why the fact was recorded so many times and correct the issue.
- To associate the specified event with a report group, click Add Report Group.
- Click the report group to add.
Note: Do not add a dimension that is populated by an event to the same event. For example, if Dimension X is populated by Event Y, do not make Dimension X a dimension of Event Y. All values and dimensions are written at the same time. In the above scenario, the value of Event Y is unpredictable.
Note: You can't add a cxOverstat report group to a non-cxOverstat event.
- When you add all wanted report groups, click the next step.
In the following example, you have a Purchase Event that records the total value of the products that were purchased. In this case, price is the event value. You could then specify a report group to record the following dimensions:
- Membership Type
- Credit Card
- Product
- Price (event value)
The recorded purchases for a day look like the following:
| Membership Type | Credit Card | Product | Price |
|---|---|---|---|
| Bronze | Amex | Plane | 72 |
| Bronze | Disc | Plane | 72 |
| Gold | Amex | Car | 24 |
| Gold | Visa | Boat | 45 |
| Gold | Visa | Car | 32 |
| Silver | Amex | Boat | 43 |
Depending on how you configured your report group, you may be able to perform additional segment analysis by examining the purchases by Gold members who used Visa cards. Because you specified that the dimensions Membership Type, Credit Card, and Product are to be recorded in the same report group with the value price, you can segment the data for further analysis. For example, you can examine just the purchases where Membership Type = Gold.
| Membership Type | Price | Credit Card | Price | Product | Price |
|---|---|---|---|---|---|
| Bronze | 72 | Amex | 72 | Plane | 72 |
| Bronze | 72 | Disc | 72 | Plane | 72 |
| Gold | 24 | Amex | 24 | car | 24 |
| Gold | 45 | Visa | 45 | boat | 45 |
| Gold | 32 | Visa | 32 | car | 32 |
| Silver | 43 | Amex | 43 | boat | 43 |
However, if you choose to record those dimensions but did not include them in the same report group, you could not see the purchases of Gold members who used Visa cards.
- You could see the purchases of Gold members, or the purchases by Visa cards, but not both together.
- Data for each report group is stored in individual tables. Since the data for two unconnected report groups is stored in different tables, there is no common key to specify how the data in one table is associated with data in the other table.