trailhead

Learning Objectives

After completing this unit, you’ll be able to:
  • Understand why the story in your data can be difficult to spot.
  • Try Einstein Discovery with a Developer Edition org.
  • Upload a dataset.
  • Fix minor data issues.

There’s a Story Buried in the Data

Every dataset tells a story. But it’s difficult to tease out that story when you’re dealing with enormous tables of data and many variables with complex relationships. Take the example we’ve been looking at, shrinking margins in an auto parts supply company. To do a thorough analysis, you consider not just inventory, but also particular markets, distributors, incentives, and likely many other factors.

Your supply company feels the pain of shrinking margins. But learning the source of that pain by sifting through a huge amount of data isn’t a pleasure drive. Luckily for us, we have Einstein Discovery. Not only does it quickly analyze huge amounts of data to explain causes, it also suggests fixes and predicts what happens if we don’t change anything.

Put yourself back behind the wheel as the VP of operations, and let’s create your first story.

Try Einstein Discovery with a Developer Edition org

Before you work through this Trailhead module, sign up for a free Analytics-enabled Developer Edition org. This org is a safe environment where you can practice the skills you’re learning.
Important

Important

For this trail, you can’t use an existing Developer Edition org. You must sign up for this special Developer Edition because it comes with a limited Analytics platform license, has an Einstein Discovery permission set license, and contains sample data required for this trail. Even if you already have an Analytics-enabled Developer Edition org, sign up for a new one now. The older Analytics-enabled Developer Edition orgs don’t get recently released features, so signing up for a new one ensures that you get the latest and greatest.

Let’s get you set up so you can log in and get started.

  1. Go to developer.salesforce.com/promotions/orgs/analytics-de.
  2. Fill out the form using an active email address. Your username must also look like an email address and be unique, but it doesn’t need to be a valid email account. For example, your username can be yourname@analyticsrocks.de, or you can put in your company name.
  3. After you fill out the form, click Sign me up. A confirmation message appears.Confirmation message appears, asking you to check your email.
  4. When you receive the activation email, open it and click the link.
  5. Complete your registration, and set your password and challenge question.
    Tip

    Tip

    Write down or remember your credentials. To log in and play, just go to login.salesforce.com.

  6. Click Save.

    You’ll be logged in to your Analytics Developer Edition org and redirected to the Setup page.

Way to go! You now have a Salesforce org with your data!

To be fair, we have made things a little easier for you in this trail by creating an org that already has all sorts of goodies ready to go. The org has Analytics and Einstein Discovery enabled, it has all necessary permissions defined, and it includes some pre-loaded test data. If you want to learn about the admin work for enabling and setting up Analytics, check out the trail called Build and Administer Analytics.

Signed up? Great! Let’s jump right in.

Download the Data

Before we can create a story, we need the data. We’ve prepared a file with the data you need to create a dataset for the auto parts company in our example. Download the CSV file called APdist.csv and save it to your computer.

Import the Dataset

The next step is to get the data from the CSV file into Einstein Discovery.
  1. In your new DE Org, click the App Launcher App Launcher icon, search for Einstein Discovery, and then click the tile to launch Einstein Discovery.
    App Launcher
    Note

    Note

    Be sure to set your popup blocker to allow popups. Otherwise, your connected app doesn’t launch.

  2. Click Datasets.Upload a dataset
  3. Click the tile labeled CSV.
  4. In the file-selection window that opens, find and select the CSV file—APdist.csv—you downloaded. Click Open (or the equivalent button in your operating system) to upload the file. Dataset after it is uploaded
    Einstein Discovery imported the data—great! But there’s a message saying “Data Improvements Detected”. Oops. That means there are some data quality issues. It’s not unusual to find all types of errors and inconsistencies in source data, but let’s clean those up because they can skew our results. Never fear—Einstein Discovery makes it easy to fix them. Click Review Changes to see a list of issues with the data.
    Note

    Note

    Don’t worry if the images here differ slightly from the screens you see in Einstein Discovery. The interface elements should be the same, but some of the details—including the data they show—may differ slightly.

    Review changes page

    The first issue seems to be a typo in a product name. We know that our auto parts company sells Ball Bearings, not Ball Bea-rings, so let’s fix that mistake.

  5. In the row showing the Ball Bea-rings issue, choose Find And Replace from the menu in the Action column.
  6. Click Add.
  7. Enter Bea-rings in the find field and Bearings in the replace field.

    You can ignore the other data issues for now. The dataset does not have to be perfect for Einstein Discovery to do its work.

    Data prep screen with find-replace of mangled data
  8. Click Apply in the upper right.
  9. Almost there! Look for the field name Margin. If Type is set to Text, click the down arrow and change it to Number. If it’s already set to Number, you’re good to go.Change the Margin field name type from text to numbers.
  10. If you made a change, click Apply in the upper right corner to apply the change.
That was easy! Remember, this is sample data. Your real data often has many more issues than that, but Einstein Discovery is excellent at finding them. Now that you’ve got the dataset in place, let’s find out what it’s trying to tell you. You’ll do that in the next unit.
Time Estimate
retargeting