12.2.2020 Training Session
Wendy Tseng | Watch Recording
Key Takeaways
- If you want to only have a description of what happened, Tableau is fine. But if you want to actually understand why things happened to then make improvements, you will need JMP.
- Tableau and JMP will most frequently co-exist
- Tableau is a reporting tool, great for executives because of aesthetics
- JMP is needed in order to design new products, processes and solve problems
- JMP Live is needed in order for scientists, engineers and analysts to seamlessly collaborate
- Ask the same questions you would to Tableau users as Excel users: How are you doing root cause analysis? Are you using designed experiments to innovate? How do you determine product reliability?
- Organizations with Tableau Server: make sure they are aware of JMP Live as a way for collaboration and sharing of statistical output
Tableau Products Compared to JMP Products:
|
Users
|
Tableau
|
JMP
|
|
Creators
|
· Tableau Prep Builder
· Tableau Desktop
· Tableau Online
|
· JMP / JMP Pro
· JMP / JMP Pro
· n/a
|
|
Consumers
|
· Tableau Data Connector
· Tableau Reader (free)
· Tableau Server
· Tableau Public
· Tableau Online
|
· JMP Live
· Interactive HTML
· JMP Live
· JMP Public
· JMP Live hosted
|
Questions You can Answer
|
Tableau
|
JMP
|
|
· What happened?
· Where did it happen?
· When did it happen?
Statistical analysis in Tableau is limited to simple linear regression and time series forecasting. Tableau integrates with R and Python which is how you can “do analytics” with Tableau.
Here are 2 resources you can look at to learn more about analytics with Tableau:
· Advanced Analytics Whitepaper
· Accelerate Your Advanced Analytics
|
· What happened?
· Where did it happen?
· When did it happen?
· Why did it happen?
· How did it happen?
How can we improve and/or optimize it?
|
** The Users: Tableau and JMP “Creators” are not usually the same people
Similarities between Tableau and JMP:
- JMP’s Graph Builder and Tableau Desktop have similar interfaces
- JMP and Tableau (Tableau Data Prep) can both prepare data and save workflows for doing so<
- JMP Live and Tableau Server both allow for non-builders to consume content
- You can hide and exclude points in visualizations (including recalculating a simple linear regression fit) using an exclusions filter (it’s like our LDF).
- Rich library of online learning (recorded and live) and paid training
- Technical support (they have SE’s, too)
What is great about JMP over Tableau
- You have the ability to make graphs in JMP AND do all the other statistical analyses you need to be able to do as a scientist / engineer
- Build models to improve or optimize products and processes
- Design Experiments to make new products / processes or improve them
- Product reliability; answer questions like:
- When will my product fail?
- How many of my products will fail by n cycles/days?
- What expiration date can I use to set my product?
- Are my processes stable and capable?
- Are my measurement systems (gauges, instruments, etc.) accurate and consistent?
- Easier to create new columns and formula columns in JMP
- JMP is a single platform for data prep and analytics
- JMP’s learning opportunities are richer (free and paid)
- JMP’s Technical Support is more attentive and hands-on
What is great about Tableau over JMP
- You have more control over the appearance of the graphs and many of the customizations are easier to do
- Visualizations are more aesthetically pleasing
- Dashboard filtering is much easier to set up in Tableau Desktop
- Setting up controls for publishing, organizing and provisioning is easier in Server
- Publishing: No scripting required
- Organizing: There are projects spaces vs. just groups
- Provisioning: You can control access to content easier due to the organizational structure
- Offers strategic, annual benchmarking assessments to customers
How can Tableau and JMP co-exist?
- I think in most situations, JMP would not replace Tableau but rather co-exist with it. Commonly customers does analysis in JMP and recreates their reports in Tableau for executive reporting.
- Tableau Desktop users are typically a different set of people than JMP users. Tableau is used more heavily in business and operations and organizations that report on metrics for monitoring something (i.e. think COVID dashboard). In our core customers, Tableau may be used for manager/executive reporting (# of defects over time, # of lab requests over time, etc.) while JMP is used for collaboration between scientists and engineers for purposes of making new products (DOE), discovery, root cause analysis.
- Where there may be some overlap is with respect to process monitoring with control limits. You do not have the ability to calculate control limits or build any control charts automatically in Tableau but scientists and engineers may choose to publish control charts in Tableau (using JMP or other tools like SQL or R to do the calculations) because of the aesthetics of the control charts and the ease of the ability to automate publishing of the charts and organize projects in the Tableau Server environment. Here is a link to an example on Tableau Public.
- In cases where our core users are using Tableau Desktop and/or Server, I think you can think of them as you would Excel users. They are missing out on data prep automation (unless they also purchased Data Prep), DOE, root cause analysis, … all the goodies that we have.
What should you listen for?
Tableau Desktop/JMP
- Any type of analysis that requires statistics. You can only do simple linear regression in Tableau (single predictor) and time series forecasting (basic stuff: exponential smoothing with additive or multiplicative model, adjusting for seasonality). However, Tableau integrates with R and Python which is how people are doing statistical analysis in Tableau.
- These are the typical questions you would be listening for. If users are only using Tableau, you can think of them as being just like Excel users.“We need to be able to…
- ..design new products / processes faster.”
- …optimize our products / processes faster.”
- …figure out the root cause of our failures.”
- …quickly identify why our products are out of specification.”
- …more efficiently prepare our data.”
- …figure out when our products will fail?”
- …determine how many of my products will fail by n cycles/days.”
- …determine if our processes are stable and capable/”
- … my measurement systems (gauges, instruments, etc.) accurate and consistent?
Tableau Server/JMP Live
- Look for customers who want to be able to collaborate between scientists and engineers more effectively with respect to statistical process control monitoring, reliability, models, root cause analysis, etc.
- Customers with Tableau Server will not be able to share any type of output that requires statistics (beyond simple linear regression and basic time series forecast graph).
|
Scenario
|
Our Opportunity
|
Guidance
|
Questions to ask
|
|
Users have Tableau* and JMP
*Desktop or Desktop+Server
|
JMP Live and/or more JMP
|
If users are not fully leveraging JMP, it would be important to focus on that to make sure they are aware of the capabilities of JMP above Tableau Desktop or that a broader user base is proficient in JMP.
If they have enough JMP users, JMP Live may be an opportunity so that they are not recreating their work in Tableau for sharing.
|
What questions are you trying to answer in Tableau? What questions are you trying to answer in JMP?
How are you collaborating and sharing outputs with other stakeholders? Did you know that we have a product called JMP Live?
|
|
Users have Tableau Desktop only (no Server)
|
JMP
|
These scientists and engineers are not able to advance their work with just Tableau Desktop. Think of them as Excel users. They can more easily prepare their data in JMP and do all the statistical analyses that is necessary to advance their work and compete.
|
• What problems are you trying to solve with Tableau?
• Are you able to identify root causes of your problems and potential solutions?
• How much time do you spend preparing your data for analysis?
|
Some Tableau Dashboards on Public
