Inline Cohort: define up to 5 cohorts (user segments) inline with every formula: Query, Funnel, Path and Retention. There is no need to jump back and forth to create cohorts any more.
Currently two cohort types are supported. More types like Retention and Funnel are coming soon in future releases.
Performed: users who performed certain events in the time window. This is the mostly used cohort type.
Query: reference a previously executed Query with condition.
Any inline cohort can be saved with a name, so they can be referenced directly in the future.
Now you can also combine one breakdown (dimension property) with cohorts, as shown in the screenshot below.
2021-12-15 R226
Cancel Formula Execution: support cancel when a formula are being executed.
Ending Event Optional in Path: support open ended path
Click on chart name to rename: click on a chart's name to rename it
2021-12-02 R224
Breakdown by Two Fields: support up to two fields in Breakdown.
Alias for Compound Measure in Query: support specifying alias for a compound measure (+ or / between simple measures)
2021-11-23 R223
Dashboard Refresh: starting this release, to speed up the loading, Dashboards will load the previously executed charts by default and only execute with latest data when the Refresh button is clicked on.
2021-11-10 R221
Negative Events in Funnel: exclude some events between steps when Order is Sequence in Funnel.
2021-10-20 R218
Funnel Order: three order options in Funnel. For three steps with events e1, e2, e3
Any: matches e1, e2, e3 in any order regardless their timestamps. Previously this was the default order.
Sequence: matches e1 -> ... -> e2 -> ... -> e3 when they happen sequentially, though there could be other events in between.
Exact: matches e1 -> e2 -> e3 in exact order without any other events in between.
2021-10-13 R217
Cohort Subject: support cohort by User, Device or Session (if available)
Cohort Time Window Between: support "between last X and Y days" in cohort time window
Q3 2021
2021-09-27 R215
Abstract Event: treat both event name and a property as a separate step in Path.
For example: "page_view: Homepage", "user_action: Checkout"
The list of abstract events and properties to use are customer specific
Session Renewal: automatically renew session when Kubit is being used
2021-09-20 R214
New feature - Effect: added support in Query to show the impact of a Trigger Event before and after the Trigger Date.
Select "Trigger" as the type for the date range
Choose the Subject to analyze by (if applicable)
Specify the details of the Trigger Event
The returned chart will be rendered with "Day 0" in the middle, then
"Day -N" on the left showing before the Trigger Date
"Day +N" on the right showing after the Trigger Date
Improvement: increase to show up to 3000 events to choose from in the Event input.
Bug fix - Cohort Pre-filter with Negate Condition: fixed a bug related to cohort pre-filter when negate condition was specified ( "count(event) = 0" ).
2021-09-16 R213
Median Function in Query: support statistic function median() in Query.
Formula Input Tuning: various UX tuning on inputs in all Formulas
2021-08-18 R210
Breakdown Bucketing: support bucketing for numeric field in breakdown. For example: breakdown by Age or Install Days with pre-defined bucket definitions.
Relative Time Window for Install + X days: when Install Date is used for group by, the analysis will only consider all the events happened within X days since Install Date.
Note: this feature is only enabled for Query in this release. Funnel will come in the next release.
2021-08-11 R209
Subject Bucketing and Histogram: Support group by Subject (user, device or session) in Query with bucketing (occurrences of Subjects in each bin) and show the distribution in histogram with statistics (min, max, mean, median, standard deviation and 95-percentile).
For example: Count Event(Download) for Subject=User and Bucketing=Range10
Query returns for each date, the count of Users who have a certain number of Download events for each bin (0-2, 2-4, 4-6, 6-8, 8-10, 10+).
In bar chart view, right click a certain date to "Expand" it to show the histogram to get a sense of the distribution of the occurrences
Expand Bar Chart: right click a certain date on bar chart to "Expand" it.
2021-07-22 R205
Retention: Support Subject. Applicable customers can now analyze retention by user or device.
Switch to Slack Channel for In-App Support. Access support in Slack from anywhere within the application.
2021-07-19 R204
Query: Time Window for Install date. When Install date is selected, the look forward time window must be specified now (number of days). This means for the users with an Install date (cohort), analyze how many days of Events. Previously this value is hard coded as 90 days.
Query: New Time Bucket - All Time. A new time bucket A (stands for "All Time") is introduced. When All Time is selected, the data within the selected date range is treated as a whole instead of grouped by Day/Week/Month/Year.
Retention
New Cohort Time Bucket - All Time. Treat all users within the selected date range as a whole instead of separate cohorts by Day/Week/Month/Year.
Retention Time Window: analyze how many days of data for each cohort. Leave it blank as a default to restrict to only the data within the selected date range (show a triangle of retention chart).
2021-07-02 R203
Increase field value limit to 25: allow up to 25 values to selected in a filter for any field
Session Duration: only for applicable customers
Cumulative Sum View: show the Cumulative Sum in line/bar chart for Query
Q2 2021
2021-06-22 R202
Up to 500 Groups in Export: increase to 500 groups when Export to CSV or Jupyter
A/B Test Experiments and Treatments/Variants: only for applicable customers
Measure and Alias Improvements:
Default optional, using + sign to specify value
Use Event name as default alias for Measure (in Query) and Step (in Funnel)
2021-06-08 R201
Subject and Partition in Funnel/Path:
Provide the option of analyzing by User or Device (if applicable) as Subject
By default, Partition the events by Day (all events have to happen within the same calendar day). For certain applicable customers, Partition by Session is also provided (all events have to happen within the same session).
2021-06-01 R200
Path Usability Improvements: make it easier to show information about an edge; wrap text for long event names
Promote Schedule to the top level menu on the side
2021-05-26 R199
Session Support: Unique Session function and First Session flag (Only for selected customers who provides required data like session_id and first_session_id)
Breakdown by Install Date, Install Week and Install Month: exact field names may vary by customers
2021-05-19 R198
Support IdP Initiated SSO: Single Sign On from authentication provider directly
Dynamic Join with Dimension Tables: only join with dimension tables when a dimension field is specified in Formula to boost performance
2021-05-12 R197
RudderStack Integration: open-source CDP
Filter by Date Fields: eg Install Date
2021-05-06 R196
Nested Property: expose the internal properties of a JSON or BigQuery Struct field for Filter, Breakdown or Function.
2021-04-28 R195
New Filter Widget
2021-04-21 R194
Performance Improvements: query optimization, mostly for BigQuery/Firebase customers
2021-04-14 R193
Reorder Funnel Steps and Query measures: use drag-n-drop to reorder the steps in Funnel or Measures in Query
Insert steps in Funnel: insert steps in the middle of the funnel
2021-04-06 R192
Copy SQL for every analysis: available for Enterprise tier or as Add-on package
Bug fix: Edit Dashboard chart order issue.
March 2021
Data Dictionary: show detailed information about Field, Event, Cohort, Measure, Filter.
UI improvements: compact sidebar; moved Existing Cohorts to Dictionary.
Search: Event, Field and Field Value search improvements.
February 2021
Named Measures: save a Measure which can be shared by others
Retention: Classic and Range retention, supporting different time bucket for Cohort and Retention
January 2021
mParticle Integration: ingest mParticle events.
Braze Integration: import cohort information to Braze to segment users with push/email campaigns.
Prediction (beta): shows how some action events can “predict” the probability which users will move from a source cohort to a target cohort.