TDX Tuesday: How to Reopen with Intelligent Predictions from Tableau and Einstein Discovery

Hey #AwesomeAdmins! Every Tuesday this month, we will talk about the awesome TDX sessions we saw last month. We realize you may not have a whole day to set aside for learning, but we bet you can find 30 minutes this week (and next week, and maybe the week after that!). ?

Admin Episode: How to Reopen with Intelligent Predictions from Tableau and Einstein Discovery

The Admin Channel at #TDX20 features an awesome session on Tableau and Einstein Discovery with John Demby. During this session, John talks about what Tableau can do when we actually look deeper and explore data. We’ll bring Einstein to the party and look at what Einstein Discovery can bring us in terms of predictions and analytics. Then, we’ll bring our old friend the Lightning Web Component in to actually show how we can make those insights accessible— all on the Salesforce Platform.

Three key takeaways:

1. Data-driven decisions must include some level of predictive intelligence. If you really want to drive those insightful decisions, you need to understand where you’re headed.
2. Agility and speed are critical, especially today, because the ability to make key business decisions can impact customers’ lives, your employees’ lives, and the community in general.
3. Insights need to be both actionable and accessible. Simply developing insights is one thing; being able to put them in front of people who can take action on them is critical.

Working in Tableau demo

In our demo example, we talk about Caroline, a Data Analyst who’s been given A LOT of new data. She takes a look at this data, explores it, and tries to give the best possible insights to leadership and all involved so that they can begin to make those critical decisions going forward.

Caroline’s been given some data from the operational team around closed stores and open stores. First, she looks at this data from a geospatial perspective, around mapping and things like that. Then, she notices the information about employee sick days and things like that. So she creates a base map including the store IDs. Now, she knows all the stores across the country that have reopened since being closed due to COVID-19. But what she really wants to do is look at this data from an employee health perspective.

Einstein Analytics and Tableau

Now it’s time to think about what we could do with this data in Einstein Analytics. More specifically in this case, Einstein Discovery, because if we bring this data into Einstein, we can use it to seed a model.

Einstein Discovery is a wizard-based simple approach. It doesn’t require you to be a Data Scientist to create a model. After going through the wizard and leaving the default settings, Einstein begins to crank through the data and look at a lot of things, like patterns and matches. It will look at different variables and try to decide which ones impact the prediction. What Einstein does, in a short span of time, would take a Data Scientist on your staff maybe weeks or even months—to build this model, test it, run it through, and test it again.

In our example, Caroline wants to minimize employee sick days using predictions, so that leadership can decide intelligently what they want to do with the different stores. Very quickly, Einstein Discovery shares great insights about the data. She sees that management training is critical and tenure is pretty important. She also learns that, in some cases, Employee Safety Training actually results in a higher number of employee sick days, which tells her the training needs to be reviewed, updated, and fine-tuned. After reviewing the insights, Caroline can also go behind the scenes and validate the model. She can see what Einstein is doing with the coefficients, and take a look at the R Code. (We don’t use R in Einstein, but we do create an R output. This enables a Data Scientist to look at that data and do more with it.) Once Caroline deploys the model, it becomes available in Tableau through the Einstein API. Now, she has a model that she can come to over and over again—she can fine-tune it and get that insightful and actionable information.

Using the power of Tableau to visualize the data and the power of Einstein to get the predictions, we can obtain the insights we, as admins, need to move forward.

Use Lightning Web Components to surface Tableau in Salesforce

Tableau with Salesforce has developed a brand new Lightning Web Component. Soon, you’ll be able to get this web component from the AppExchange and use it like any other component in Lightning App Builder. This way, important data is surfaced in the workspace of the user who needs it. If the information gets updated in Tableau, it’ll automatically be updated here. Anybody making decisions can see this information and interact with it.

What’s next? Take these three steps

First, we would love for you to have a conversation. A conversation with us here at Salesforce, Tableau, and Einstein. And, a conversation with your leadership, with those who curate and understand all of your data inside your organization. Start talking to start understanding.

Next, start using Tableau. You can download it and start using it with a 14-day trial. Or you can use Tableau Public, which is really the YouTube of data visualization with more than two million visualizations already there! It’s the perfect place for you to start your journey.

And finally, we want you to be the Data Superhero of your company. Data is a very powerful thing. When we use it, and use it properly, we can change the course of history. We can change not only the course of our company’s success but also the course of our careers. So many admins have literally made a career out of being a Data Superhero!

Resources for #AwesomeAdmins

Coming up tomorrow, on Trailhead Live, check out our session for #AwesomeAdmins: Explore the Power of Data with Tableau.