Recently, I worked on a project for a client using Tableau's Admin Insights. The documentation provided by Tableau is sparse, so I had to experiment and investigate in order to understand the data sources I was working with. In particular, I worked with the TS Events, TS Users, and Site Content data sources included in Admin Insights. Here is a summary of what I've learned, including data dictionaries for TS Events and TS Users and overviews of all three datasets.
Image created by ChatGPT
Admin Insights Overview
Admin Insights is a Tableau Cloud project that includes curated data sources and pre-populated workbooks using your site's data. Here, we are focusing on the data sources so we can use them to create our own custom workbooks that meet our needs.
In general, the Admin Insights data sources show data from the last 90 days, although exactly what this means can differ depending on the data source and field.
For example, in TS Events, each row of data is one event performed by one user at a specific datetime (Cecily Santiago Published a Workbook on 2024-11-06 13:52:45), so this data source represents all the events that occurred on the Tableau Cloud Site in the 90 days before the data was pulled.
However, the TS Users table provides a snapshot of all the users on the Tableau Cloud Site at the time the data was pulled and stats about them aggregated over the past 90 days, although some of the fields may contain data from more than 90 days ago. Here are some examples of fields and how the date comes into play:
“User Creation Date” will remain populated and static, even if the user was created more than 90 days ago.
“Last Login Date” will show the most recent time the user logged in to the site as of when the data was pulled even if that login was more than 90 days ago.
“View - Last Access Date” will show the most recent time the user accessed a View if they accessed a view in the last 90 days. Otherwise, it will show NULL, whether they ever accessed a view or not.
“Access Events - Views” will show the cumulative number of times the user accessed a view in the last 90 days.
Now let’s dive into each of these three datasets. What I’m presenting is my best guess at a data dictionary. I investigated to confirm my understanding when possible, but there could still be errors. Due to the scope of my project, I am able to present a more detailed understanding of TS Events and TS Users than of Site Content.
TS Events
In this dataset, one row of data represents one event. The event types are Access, Create, Delete, Publish, Send E-Mail, and Update and an event occurs when an Actor (site user) performs one of these actions. Each event has a datetime. The unique identifier in this table is “Event ID,” which is a ten digit number.
Joining TS Events to TS Users and Site Content
These are the most common joins you would perform, although others are possible depending on what questions you want to answer.
Join TS Events to TS Users on "Actor User ID = User ID" to bring in details of the Actor for each event.
Join TS Events to Site Content on "Item ID = Item ID" AND "Item Type = Item Type" to get details of the item being acted upon in each event.
TS Events Data Dictionary
Please note that I have created this data dictionary based on my best understanding of the data. I have verified field meanings when I was able to, but have not checked everything. There may be errors or misinterpretations. I have included a "Confidence Level" for each field to show how sure of my interpretation I am. If you find an error, please reach out to me so I can correct it.
Field Name
Data Type
Meaning
Notes
Confidence Level
Site Luid
String
Unique identifier for the tableau cloud site - only matters if you have more than one site
The same in each row
High
Historical Project Name
String
The original name of the project that holds the item being acted upon when the project was first created
Moderate
Site name
String
Name of your tableau cloud site
The same in each row
High
Actor User Id
Integer
The User ID of the user doing the action
High
Target User Id
Integer
If the action affected another user (i.e., Delete User), this is the User Id of that user
High
Actor User Name
String
The User Name of the user doing the action
High
Event Name
String
The type of action and what type of object it is acting on
Examples: Update Flow, Publish View, Access Data Source
High
Event Type
String
The six types of actions
All Possible Values: Access, Create, Delete, Publish, Send E-Mail, Update
High
Event Id
Integer
The unique identifier for rows in this table
High
Event Date
DateTime
DateTime of the event in UTC
High
Actor License Role
String
The type of license of the Actor
All Possible Values: Creator, Explorer, Viewer
High
Actor Site Role
String
The assigned site role of the Actor
All Possible Values: Creator, Explorer, Explorer (can publish), Site Administrator Creator, Unlicensed, Viewer
High
Item Type
String
The type of item the action is acting on. It is NULL when the action acts on a user, project, flow, etc.
All possible values: Null, Data Source, View, Workbook
High
Item Id
Integer
The id of the item being acted on
High
Item LUID
String
The "Locally Unique ID" of the item being acted on
High
Item Name
String
The name of the item being acted on
High
Workbook Name
String
The name of the workbook the item being acted on is in, if applicable
High
Historical Item Name
String
The original name of the item being acted upon when it was first created
Moderate
Project Name
String
The name of the project where the item being acted upon is stored
High
Item Owner Id
Integer
The user id of the user who owns the item being acted upon
High
Item Owner Email
String
The email address of the user who owns the item being acted upon
High
Item Repository URL
String
Low
Historical Item Repository URL
String
Low
Admin Insights Published At
DateTime
DateTime this data was pulled
Seems to be local time - unsure
High
Count of Distinct Actors
Integer
Number of actors doing the action
It’s always 1. Probably there to be used for counts
Moderate
Event Date (local)
DateTime
Local DateTime the event occurred
Local to the user or where the cloud site is hosted?
High
TS Users
In this dataset, one row of data represents one current user of the Tableau Cloud Site. This includes “Unlicensed” users but not deleted users. A user who owns content cannot be deleted, but they can be unlicensed, which means they will remain a listed user but will not have access to the site. “User ID” or “User LUID” can be used as unique identifiers in this table. This is a snapshot of the users and their info from the past 90 days.
This dataset also contains some global information about the site that is not user specific–various breakdowns of number of licenses at the time the data is pulled, the site version at the time the data is pulled. This data will be the same in every row.
Joining TS Users to TS Events and Site Content
These are the most common joins you would perform, although others are possible depending on what questions you want to answer.
Join TS Users to TS Events on “User ID = Actor User Id” to get the events for each user.
Join TS Users to Site Content on “User Email = Owner Email” or “User Email = Item Parent Project Owner Email” to get the site content each user owns.
TS Users Data Dictionary
Please note that I have created this data dictionary based on my best understanding of the data. I have verified field meanings when I was able to, but have not checked everything. There may be errors or misinterpretations. I have included a "Confidence Level" for each field to show how sure of my interpretation I am. If you find an error, please reach out to me so I can correct it.
Field Name
Data Type
Meaning
Notes
Confidence Level
Site Luid
String
Unique identifier for the tableau cloud site - matters if you have more than one site
The same in each row
High
Site Name
String
Name of your tableau cloud site - matters if you have more than one site
The same in each row
High
Allowed Creators
Integer
Number of creator licenses you have access to
The same in each row - this includes used and unused licenses
High
Allowed Explorers
Integer
Number of explorer licenses you have access to
The same in each row - this includes used and unused licenses
High
Total allowed licenses
Integer
Overall number of licenses you have access to
The same in each row - this includes used and unused licenses
High
Allowed Viewers
Integer
Overall number of viewer licenses you have access to
The same in each row - this includes used and unused licenses
High
User ID
Integer
A unique identifier for users.
High
User LUID
String
Another unique identifier for users.
“Locally unique identifier”
High
User Email
String
The email address of the user.
High
User Name
String
The user name of the user
Coincides with User Email but may have different capitalization
High
User Friendly Name
String
Last Name, First Name
High
User Creation Date
DateTime
DateTime the user was added to this site - can be further back than 90 days
Unsure if UTC or Local time zone
High
User Account Age
Integer
Number of days from User Creation Date until the data was pulled
High
Last Login Date
DateTime
Date time of last login - can be further back than 90 days
Null means never logged in
High
Days since last login
Integer
Number of days from Last Login Date until the data was pulled
Null means never logged in
High
User license type
String
The license type the user is assigned as of when the data is pulled
Creator, Explorer, Unlicensed, Viewer
Unlicensed means they no longer have a license
High
User Site Role
String
The site role the user is assigned as of when the data is pulled
Creator, Explorer, Explorer (can publish), Site Administrator Creator, Unlicensed, Viewer
Unlicensed means they no longer have a license
High
Projects
Integer
The number of projects the user owns as of when the data is pulled
Moderate
Data Sources
Integer
The number of data sources the user owns as of when the data is pulled
Moderate
Certified Data Sources
Integer
The number of certified data sources the user owns as of when the data is pulled
Moderate
Size of data sources (MB)
Decimal
Total MB for all data sources the user owns as of when the data is pulled
Moderate
Workbooks
Integer
Number of workbooks the user owns as of when the data is pulled
Moderate
Size of workbooks (MB)
Decimal
Total MB for all workbooks the user owns as of when the data is pulled
Moderate
Views
Integer
Total number of Views the user owns as of when the data is pulled
Moderate
Access Events - Data Sources
Integer
Total number of times the user has accessed a data source in the 90 days before the data was pulled
High
Access Events - Views
Integer
Total number of times the user has accessed a view in the 90 days before the data was pulled
High
Publish Events - Data Sources
Integer
Total number of times the user has published a data source in the 90 days before the data was pulled
High
Publish Events - Workbooks
Integer
Total number of times the user has published a workbook in the 90 days before the data was pulled
High
Data Source - Last Access Date
DateTime
Last time the user accessed a data source in the 90 days before the data was pulled
NULL means they have not accessed a data source in the 90 days before, but may have done so previously.
This is probably in UTC
High
Data Source - Last Publish Date
DateTime
Last time the user published a data source in the 90 days before the data was pulled
NULL means they have not published a data source in the 90 days before, but may have done so previously.
This is probably in UTC
High
View - Last Access Date
DateTime
Last time the user accessed a view in the 90 days before the data was pulled
NULL means they have not accessed a view in the 90 days before, but may have done so previously.
This is probably in UTC
High
Workbook - Last Publish Date
DateTime
Last time the user published a workbook in the 90 days before the data was pulled
NULL means they have not published a workbook in the 90 days before, but may have done so previously.
This is probably in UTC
High
Web Authoring - Last Access Date
DateTime
Last time the user used Web Authoring in the 90 days before the data was pulled (UTC)
NULL means they have not used Web Authoring in the 90 days before, but may have done so previously.
This is probably in UTC
High
Total Traffic - Data Sources
Integer
Number of visits to data source this user owns - 90 days or all time?
Low
Unique Visitors - Data Sources
Integer
Number of users who visited data sources this user owns - 90 days or all time?
Low
Total Traffic - Views
Integer
Number of visits to Views this user owns - 90 days or all time?
Low
Unique Visitors - Views
Integer
Number of users who visited a View this user owns - 90 days or all time?
Low
Admin Insights Published At
DateTime
DateTime the data was pulled.
Can be used as “As Of” date. Seems to be in Local Time
High
Tableau Desktop - Last Access Date
DateTime
DateTime this user last accessed Tableau Desktop, even if more than 90 days ago. (UTC)
Unsure how it tracks desktop access
Low
Tableau Desktop - Last Product Version
DateTime
Last version of Tableau Desktop this user downloaded or published from or signed into?
Unsure how it tracks this
Low
Tableau Prep - Last Access Date
DateTime
DateTime this user last accessed Tableau Prep, even if more than 90 days ago. (UTC)
Unsure how it tracks desktop access
Low
Tableau Prep - Last product version
String
Last version of Tableau Prep this user downloaded or published from or signed into?
Unsure how it tracks this
Low
Pulse Metrics
Integer
Number of Pulse Metrics this user owns
Moderate
Site Version
String
Version of Tableau this site is on
Not user specific - the same in all rows
High
Total occupied licenses
Integer
The number of licenses in use as of when the data was pulled
Not user specific - the same in all rows
High
Total remaining licenses
Integer
The number of available licenses as of when the data was pulled
Not user specific - the same in all rows
High
Occupied explorer licenses
Integer
The number of "Explorer" licenses in use as of when the data was pulled
Not user specific - the same in all rows
High
Occupied Viewer Licenses
Integer
The number of "Viewer" licenses in use as of when the data was pulled
Not user specific - the same in all rows
High
Occupied Creator Licenses
Integer
The number of "Creator" licenses in use as of when the data was pulled
Not user specific - the same in all rows
High
Tableau Desktop - Last Access Date (Local)
DateTime
DateTime this user last accessed Tableau Desktop, even if more than 90 days ago. (Local time)
Unsure how it tracks desktop access
Low
Tableau Prep - Last Access Date (Local)
DateTime
DateTime this user last accessed Tableau Prep, even if more than 90 days ago. (Local Time)
Unsure how it tracks desktop access
Low
Web Authoring - Last Access Date (Local)
DateTime
DateTime this user last accessed Web Authoring in the last 90 days. (Local Time)
High
Site Content
In this dataset, one row of data represents one item on the Tableau Cloud site. An item can be a Datasource, Flow, Project, View, or Workbook. This is a snapshot of the site content and its info from the past 90 days. “Last Accessed At” (UTC and Local) fields go back 90 days but other date fields, such as “Updated At” and “Last Published At” (both have UTC and Local time zone versions) go back to all time. This data source also gives information about the parent project and top level project for each item.
This datasource in important to use when evaluating stale content because it includes all items on the site, whereas TS Events only includes items acted upon in the 90 days before the data was pulled. If content is not being used, it will not appear in TS Events, but will appear in Site Content.
Due to scope and time restraints, I did not investigate this data source as deeply as TS Events and TS Users. Therefore, I have not created a data dictionary for this data source.
Joining Site Content to TS Events and TS Users
These are the most common joins you would perform, although others are possible depending on what questions you want to answer.
Join Site Content to TS Events on “Item ID = Item ID” and “Item Type = Item Type” to get events related to each item.
Join Site Content to TS Users on “Owner Email = User Email” to bring in info about the owner of each item.
