Use filters
This article applies to the following Customer Insights roles: Viewer; Developer
When you design a Look or a Dashboard, you have the option of including filters; filters provide a way for you to “slice and dice” the data to your heart’s content. For example, the Events Summary Dashboard lets you filter on seven different fields:
- App ID
- App Name
- Event Type
- API Client Name
- Entity Type
- Provider (IDP)
- Date Range
If you want to take a different peek at the data (for example, if you want to see data for the last 30 days rather than the last 90 days), click Filters to expand the Filters section of the Dashboard:
What if you don’t see an expand/collapse Filters link? That means that the Dashboard/Look designer didn’t include any filters.
Each filter on the Look or Dashboard includes three sections: the field name (e.g., App ID), the operator type (for example, is equal to) and the target filter value. Unless you are in edit mode, you cannot delete any of the fields, nor can you add new fields. What you can do, however, is:
- Change the operator
- Change the target value
- Add an AND or OR clause
Let’s walk through an example to show you exactly how filters work. To begin with, take a look at the Profiles By Creation Month (Last 12 Months) visualization:
As you can see, this visualization graphs the total number profiles of created each month for the past year. But suppose we’re only interested in those months when profile creations reached 300,000 or more. If that’s the case, then we can add this filter:
We’ve done two things with our new filter. First, we clicked the operator field and changed the operator to is greater than or equal to. Why? You got it: the previous operator (is equal to) would have shown us only those days where the event count was exactly 300,000. Now, we’ll see all the days where the event count was equal to 300,000 or more.
Second, we entered our target filter value: 3000000. And yes, because this is a numeric field, we can’t include commas; entering a value of 300,000 would result in an error:
A quick aside: You often don’t have to type in a complete value when configuring a filter. For example, if you click in an Event**Type**box and wait a second or two, a dropdown list appears showing all the event types currently in the database:
Instead of typing signin.traditonal_register, you can simply select the value from the list.
Of course, sometimes there are too many possibilities to fit in a dropdown list; for example, you might have literally thousands of cities in your database. In a case like that one then, as soon as the drop down appears, just start typing the city name (e.g., Albuquerque). Customer Insights will begin filtering the filter and, the desired name will eventually show up.
But back to our example. If we click Run to execute the query, we should see a marked difference in our visualization:
Needless to say, that was the whole idea behind adding the filter.
As noted, you can’t add fields to filter on; that’s up to the person who created the Look or Dashboard. However, you can filter on the same field in multiple ways. For example, suppose we’re interested in outliers: months when the event count reached or exceeded 325,000 or months when the event count failed to reach at least 250,000. This filter will do that for us:
In turn, we get a visualization and that shows the highs and the lows, and nothing in between:
To add a clause to a filter, click the + sign next to the filter of interest; that adds a second clause that filters on the same field:
Note that Customer Insights analyzes the field and its datatype, and automatically chooses the Boolean operator (AND or OR) to apply to the clause. In this case, OR makes the most sense, so the new clause is added as an OR clause.
Why does it make the most sense? Because a single month can’t have a profile count that is simultaneously greater than 325,000 and less than 250,000.
After clicking +, we selected the appropriate operator (is less than), added the search value (2500000), and then clicked Run.
If we wanted to add another clause, we can do that by clicking + again. Alternatively, we can delete either of our clauses by clicking the X next to the clause:
However, you can’t get rid of both clauses: as soon as you’re down to a single clause, the X disappears from the lone remaining clause.
And you’re right: sometimes Customer Insights might insert an AND clause when you really wanted to do an OR search: for example, you wanted to find all the users who live in Oregon or in Washington. Interestingly enough, you can do that sort of search without having to insert an additional clause of any type. Instead:
-
Click in the Region field and select OR from the dropdown list. Oregon will appear in a little gray box:
-
Click in the Region field again (in a blank spot to the right of OR) and select WA. Both OR and WA will appear in the Region field:
If you run your query now, users who live in either Oregon or Washington are returned:
And you’re not limited to just two selections; if you want to, just keep clicking and just keep adding states:
To remove a state from the list, just click the little X next to the state name.
If you want to reset the Dashboard to its original settings, but can’t remember what those original settings actually were, just leave the page and then come back.
If you’re working with a Look rather than a Dashboard, you don’t have to leave the page. Instead, just click Reset Look:
Updated about 2 years ago