Tableau Positioning

Question 1

Tableau’s Product Development team has taken the following philosophical points about metadata:

Answer 1

Analysis must be possible without first undertaking a metadata modeling exercise.

Metadata in existing systems should be leveraged when beneficial.

Metadata is a useful abstraction, but it should not be constraining.

Metadata are defaults that can be changed at runtime.

Business users don’t need to understand metadata to be successful.

Question 2

Tableau’s metadata model has 2 layers of abstraction and a run-time model that allow simplicity, flexibility and reusability.

Answer 2

Connection. Data Model. VizQL Model.

Question 3

Connection (metadata model)

Answer 3

Server
Connection Attributes
Tables
Joins

Question 4

Data Model (metadata model)

Answer 4

Defaults
Comments/Descriptions
Calculations
Aliases
User Created Fields

Question 5

VizQL Model (metadata model)

Answer 5

Filters
Aggregations
Blends
Roles
Table Calculations

Question 6

Customer Satisfaction

Answer 6

Customers remain extremely happy with Tableau (for the fourth year in a row). This was a key factor in having the highest ability to execute score in the entire Gartner survey.

Question 7

Ease of Use

Answer 7

73% selected Tableau’s product for ease of use, which place it among the top two vendors in the Gartner survey.

Question 8

Low Cost to Implement

Answer 8

Tableau customers reported one of the lowest implementation costs per user of any vendor in the survey.

Question 9

Product Score (Gartner)

Answer 9

Tableau earned one of the top aggregate, weighted average product scores (even when adjusted for enterprise deployments).

Question 10

Product Features Score (Gartner)

Answer 10

Tableau ranked 1 or 2 in the Gartner survey in the following categories:

Dashboards
Business User Data Mashup
Geospatial Intelligence
Mobile
Interactive Visualization Capabilities

Question 11

Mobile

Answer 11

Tableau has one of the highest percentages of users actively deploying mobile BI in the Gartner survey. Because it’s so easy, author once view anywhere.

Question 12

Data Volume

Answer 12

Customers report average deployment sizes in terms of users, with data volumes among the best in the Gartner survey as a result of Tableau’s direct query access.

Question 13

In-Memory Data Engine (Gartner)

Answer 13

Can be used as an alternative to, or in, hyrbrid mode with its direct query access, enables fast performance on large and multisource datasets and on complex queries, such as very large multidimensional filters.

Question 14

Migration

Answer 14

Despite frequent new product releases requiring customer to upgrade, Tableau customers gave it one of the lowest scores for migration complexity (lower means easier); over 90% of its surveyed customers were on the latest release of the software.

Question 15

What does the business intelligence landscape look like?

Answer 15

Tableau is witnessing in the business intelligence market a broad, market shift towards Tableau’s way of doing things. If you look at our financial results, I hope by now this is self-evident, but I just want to make sure it is spoken – Tableau is bringing easy, visual, affordable, analytics products to the business intelligence market in a way that has never before been achieved.

The business intelligence market today, if you just look at the products customers can go buy, the business intelligence products that currently dominate the industry are complicated and difficult to use, and slow-moving, and inflexible, and expensive. And I mean every one of those words. And into that frustrating customer landscape, Tableau is bringing an easy for everyone, affordable, fast-moving alternative. That is what we’re doing.

Question 16

What types of customers have been buying Tableau?

Answer 16

Tableau has witnessed a dramatic shift in the business intelligence market toward’s Tableau’s way of doing things. Customers have shown a tremendous appetite for easy for everyone, affordable, fast-moving alternatives to the difficult to use, slow-moving and inflexible products that the market traditionally offers.

So given that, the large deals that we are closing are happening with customers of all shapes and sizes. They’re occurring with Fortune 500 companies, they’re occurring with universities, they’re occurring with non-profits. They’re occurring with government agencies, they’re occurring with think tanks and consultancies. They world is thirsty for products that help them see and understand the data they have in their databases and spreadsheets, and Tableau is the first mass-market product available, and as a result, we’re seeing large deal products in all of our segments.

Question 17

Why do people buy Tableau in traditional cases?

Answer 17

Well let me tell you what they have in common. And then I’ll tell you what I think are the reasons people buy, and how they differ from one to another. What they have in common is that most customers out there just picture a Fortune 500 division, or a regional manufacturer, or a university or a government agency or a nonprofit, or a small business for that matter.

Just picture an organization. It has some data stored in databases, and in spreadsheets, and in places like salesforce.com and NetSuite. And they typically want to answer questions using that data. They want to answer questions, they want to make better decisions, they want to roll out executive dashboards, and roll out faster and more interactive reports, and they want to produce visualizations that can be used to persuade and inform people of what’s going on.

In short, they want to use facts to create data-driven informed decisions for their organization. That’s what they all have in common.

Now I think to the spirit of your question, compared to the alternative products that have traditionally been available in this business analytics marketplace, some customers adopt Tableau primarily because it is 10 times faster than the traditional product. And that’s just why they buy. They say, “hey look. I can complete my data dashboard project in six months using the alternative, or I can do it in six days using Tableau.” It’s literally that dramatic a difference between how fast you can do things, and that’s why some customers buy.

There’s another reason people buy. Which is a lot of these buyers that I describe also want to put data into the hands of more of their people, and these traditional systems, these traditional, enterprise business intelligence platforms that have been available in the market, they are so complicated to use and so expensive that only a small priesthood of people use them within the company. And yet most executives who lead these companies, going all the way to the presidents of companies and universities – they’re saying “look, I want data and business analytics in the hands of all of my people.”

I want everyone consulting facts before they make decisions. Not a few. And so compared to these traditional alternatives, Tableau is a self-service business analytics product. It enables companies to take business analytics out of just the IT function and put it into the hands of all the decision makers in the company, and that’s a second reason people buy, and of course some buyers are a combination of both.

Question 18

Best practices for custom SQL statements?

Answer 18

When connecting to multiple tables, choose to build a multi-table join using the regular multi-table Tableau interface, and avoid using custom SQL. Custom SQL will take the entire SQL statement in the custom SQL dialogue and “wrap” all of your human drag and drop actions inside of that statement. This might result in an onerous and complex query being sent to the database. Custom SQL can be very useful in specific situations; However, it should only be considered an option when regular multi-table joins are insufficient.

Question 19

Best practices for database driver support?

Answer 19

Always use the database drivers provided by Tableau Software, located on our support pages.

Question 20

Best practices for general order of database connectivity support?

Answer 20

The general best practice of support for databases and their drivers is:

1. Use the “First Class” database connection if it is listed in Tableau’s “connect to data” dialogue
2. Use a generic “ODBC driver” connection if #1 is not available, and
3. Use a generic “ODBC DSN” connection if #2 is not available or not working.

Question 21

Best practices for ODBC connections?

Answer 21

If possible, try to build your generic ODBC connections directly inside of Tableau Desktop, as opposed to leveraging a DSN entry in Microsoft Windows. By doing so you will be providing maximum portability when publishing from Tableau Desktop to Tableau Server.

Question 22

Best practices for managing slow connections?

Answer 22

If your live connection to your data source is slower than desired, you may want to consider a partial or full extraction of that data using Tableau’s Data Engine.

Another approach to solving a slow performance data connection is to analyze the queries that Tableau is sending to your data source with an eye for tuning that data source. These queries can be found by enabling Performance recordings directly inside of Tableau Desktop (from the help menu).

Question 23

Overview of metadata management?

Answer 23

Tableau Desktop and Tableau Server provide a wealth of features around metadata management. We have consistently been ranked as one of the top analytics companies around this area of functionality. A great place to start learning about these features is our white paper on the Tableau Metadata Model. Some of the specific tools in our toolkit that help manage metadata include:

• Complete separation of the data connection from the meta data management and experience.
• Built-in ability to recognize data types such as dates, strings, numbers, and geographies, as well as the ability to easily cast or override these choices.
• Built-in metadata features such as Tableau Groups, Sets, Bins, Hierarchies and similar.
• Built-in non-aggregate or aggregate calculation engine.
• Built-in post-database-operative aggregation calculations called “Tableau Table Calculations” which allow you to perform 2nd,3rd, or Nth pass calculations on your answer sets.
• Field Obfuscation (hide fields), field renaming (edit column names locally), and field value renaming (edit aliases)
• Single Version of the Truth controls using Tableau Server and the Tableau Server “Data Server”

Question 24

Best practices for metadata management?

Answer 24

Whenever possible, use Tableau Server and its “Data Server” component to centralize all metadata management.

Question 25

Best practices for PHI, HIPPA or Field Security?

Answer 25

There are a lot of ways our customers go about deploying Tableau to meet HIPPA requirements. Having said that, a lot of the requirements are around business process and not all the requirements apply to our software because Tableau often times is deployed simply to read the data and not to store it.

It is by design we don’t have any documentation on our website because HIPPA requirements can mean different things to different organizations. However, there are some great community threads debating how best to go about deploying Tableau with various HIPPA requirements.

One of our most successful customers I think did a great job summarizing what worked for their organization in the above thread:

“We went back and forth about where to put the patient name & date of birth in this structure, and decided to put it in the pool of data available to all quality staff a) because they already have access in our EMRs, b) we have a dual-identification policy that is needed to compare records and when directly conferring with patients, and c) it makes life easier for quality staff.

In terms of how this influences our use of Tableau, the Tableau Desktop users get to see everything they have access to, when we publish dashboards we set the permissions to the right level. Since Tableau Desktop and Server both log every query, if we ever had to we could see what someone did and what data they pulled from the system.”

The best next steps are probably to get a product consultant on the phone to prescribe the ways this customer can go about “locking down” the data with Tableau and how to audit activity as well. Ultimately, the customer has to decide on their own business processes with the tools we give them. The important thing to convey is that all our customers have the security and audit tools in Tableau to craft their own BI business processes to meet even the most stringent security standards and we have many Healthcare customers using Tableau with great success.

Question 26

Best practices on metadata feature support?

Answer 26

If desired, create common metadata objects like sets, groups, calculated fields and hierarchies directly inside the Tableau experience, as opposed to in the database directly. Two important results of doing so are: 1. Improved database maintenance and performance, 2. Centralization of the metadata experience in the Tableau layer. This is a subject which generates many opinions. Some folks think that all data definitions should live in the database, and other folks think that they should live in Tableau. This is a choice you get to make – not a hard fast rule.

Question 27

Best practices on non-aggregate calculations?

Answer 27

Materialize non-aggregate calculations (including groups, sets and bins) using the Tableau Data Engine: When connected to databases live, any non-aggregate calculations you have created in Tableau Desktop will be constructed and cached as you use the software. Running a data extraction on the source data will materialize these as true data fields for future use which will improve performance.

Question 28

Best practices on security projects and security groups?

Answer 28

The Tableau Server authorization scheme (who can do what to which content) is designed to be straightforward and elegant. The list of actions you can take upon content is deliberately small.

Tableau recommends using “groups” as opposed to individual users when assigning permissions to projects, workbooks or views. In this manner, you will not have to add or remove individuals to projects over time. Instead, you will only have to add or remove users from groups. This works whether you have integrated with Active Directory or are using Tableau’s built-in authentication.

Question 29

Best practices for testing content before moving to production?

Answer 29

Make a staging area for content. Tableau customers routinely use staging projects to park content which is still under development. Once “approved” by whatever human process is in place, it’s trivial to move content. Once you move content from Project A to Project B, you will want to reassign Project B’s permissions for that workbook. By design, this step is not performed automatically by Tableau Server

Question 30

Best practices on security settings when publishing?

Answer 30

By design, any Tableau Desktop publisher has full control over “their content”. They can decide unilaterally who can and cannot interact with their content. Meanwhile, IT organizations will have set up convenient default authorization schemas inside of Tableau Server. A best-practice step is to educate desktop publishers to leave the security settings along and then make no changes.

The Tableau Desktop publisher simply chooses the security “project” for publishing and Tableau Server takes care of all the correct permission assignments automatically.

In strict environments where an organization cannot allow desktop publishing due to security concerns, use Tableau Server’s “tabcmd” utility to publish.

Question 31

Best practices for data-driven security?

Answer 31

Tableau supports row-level and column-level security with either a manually managed process directly inside of the Tableau platform, or by leveraging data table -driven models. A common scenario is for regional managers to only see data for their regions or for doctors to only see data for their patients.

If you have no data-driven security tables, use Tableau’s built-in row-level security.

If you do have data-driven security tables, use Tableau’s built-in automatic security via calculated fields.
This method derives the access to rows of data because you have defined this access information in a security table and have then joined this table into your main fact and dimension tables.

Tableau Server Data Source Filters
Starting with version 8, Tableau Desktop and Server include a powerful new feature called “Data Source Filters”. These are hidden filters which will be applied prior to any other filter on the filter shelf.