ℹ️Objective: Use Natural Language Search (Search to Insight) to translate your questions into visualizations and understand how to run a Smart Discovery within SAP Analytics Cloud. Develop an understanding and purpose of the various pages that are created by Smart Discovery.
Estimated Time: 15 mins
Exercise Description: After hearing back from your colleagues about the dashboard you shared, you have a few quick questions regarding the company’s financial. You can quickly find answers for your questions via Search to Insight, leveraging SAC’s natural language processing capabilities. Once you try Search to Insight for yourself, you can tell your colleagues all about it and they will no longer need to come to you with their questions about data.
Once SAP Analytics Cloud found the answer to your questions you want to use other smart features such smart insight and time series forecast to enhance the result. Finally, you want to use Smart Discovery to have SAP Analytics Cloud investigate Gross Margin at BestRun and automatically built a story for you.
Key Features:
- Use Search to Insight to find answers for your questions with natural language
- Have SAP Analytics Cloud investigate one of the measures and build a dashboard for you with Smart Discovery
- Understand how Smart Insights can help you explain more of your data
- Click Home icon
ℹ️Exercise Check! Are you on the homepage?
If not, then navigate to the homepage.
🚩First you want to find out more about your Gross Margin, but you don't really know which model Gross Margin contains. As you start typing the search goes across all models and suggests you the best results for measure, dimension as well as dimension members related to what you're asking for.
- Click the Search to Insight Textbox to start asking your data related questions
-
Enter "Show me gro" in the textbox
-
Click show me Gross Margin
-
Click Search icon
ℹ️Welcome to Search to Insight (Natural Language Search)!
Drive conversational artificial intelligence forward by asking questions about your data as simple as if you were asking your colleague. Understanding natural language, Search to Insight allows you to get the information you need, instantly.
ℹ️The Search to Insight page is now providing you a chart showing you the value for Gross Margin.
🚩You can see which model the Gross Margin comes from my selecting the model icon of the chart details.
- Click Model icon to confirm the model name "ANA260_ORDER_FINANCE"
ℹ️You now want to know more about the Gross Margin per Workout.
- Click at the end of the search box
🚩You can see that several suggestions for questions are provided as part of the type ahead.
- Select Show me Gross Margin by Workout Usage (If not visible, start typing until you see the suggestion)
- Click Search icon
🚩The chart now shows the breakdown of Gross Margin by Workout Usage.
-
Click Copy icon
-
Click Copy
-
Click Exit Search to Insight
- Click Create Your First Story
- Click Responsive
-
Click Copy
-
Click Paste
🚩The chart was added to the story.
🚩You have more questions about your data. Move back to Search to Insight.
- Click the Search to Insight icon
🚩Next you want to know how the Gross Margin performs over time.
-
Type "Show me gross" in the textbox
-
Click show me Gross Margin
-
Add "by month" at the end of the search term
-
Click the Search icon
🚩You can see that Search to Insight recognized the time reference and automatically generated a time series chart on month granularity. However, you only want to look at the last 2 years. Let's adjust the question.
-
Add "for last year and this year" to the question
-
Click the Search icon
- Click Display as Line to create a line chart
🚩The new chart has been filtered on the last year and this year. You can see this on the filter token.
Note, that your data may vary based on the current year.
-
Click Filter Token to see Filter values
-
Click outside the filter box to close the menu
🚩You wonder how the business performed compared to the previous periods. Search to Insight also allows to interact with the charts directly. You can for example add recommended comparisons.
-
Right Click on the Chart
-
Click Compare To
-
Click Previous Period
🚩The variance was added providing additional insights.
🚩Let's copy the chart to the story.
-
Click Copy icon
-
Click Copy To
-
Click Page 1
-
Click Exit Search to Insight
🚩You can see the chart was also added to the page.
-
Click empty Right Lane
-
Click Menu icon
-
Click Remove
- Let's resize the chart to make it more readable
🚩You wonder how Gross Margin will perform in future. SAP Analytics Cloud can run automatic time series forecasts to provide you with this useful information.
-
Click Menu icon
-
Click Add
-
Click Forecast
-
Click Automatic Forecast
🚩You wonder what other insights SAP Analytics Cloud can provide. Specifically, you're interested in finding out more about the Gross Margin in relation to the Workout Usage. Let's add some Smart Insight to the bar chart.
-
Select the Gross Margin per Workout Usage Chart
-
Click Menu icon
-
Click Add Smart Insights
ℹ️Welcome to Smart Insights! in the chart footer.
Smart Insights lets you gain more intrinsic information about a data the data in your visualization. As you apply filters, your data updates the insights are being re-evaluated keeping your insights context aware.
🚩Smart Insights show us how Gross Margin for High Intensity workout usage has been performing overtime. Let's find out more.
- Click View More
ℹ️Welcome to Smart Insights! side panel
The side panel appears visual insights on that chart or data point.
- The upper pane shows how the information has changed over time.
- The lower pane shows information that highlights contributing factors for the data
Did you know Smart Insights can also be triggered by user who do not have edit right to a story?
Note: different charts and data may provide different insights.
ℹ️Smart Insight has inspected the different levels of the time hierarchy automatically and selected the level with the most interesting change, in this case the day granularity. However, you're more interested in the monthly data change. Switch to month view.
- Click M
🚩You can now see the change per month.
🚩Let's have a look at the top contributors
- Scroll to the bottom
🚩The data model contains 2 identical location dimensions and 2 identical dimensions containing Order IDs. This is a common assurance in business data but can result in redundant insights. A story designer can configure whether certain dimensions will be excluded from Smart Insight to provide better results. Let us exclude the redundant dimensions.
- Click Close
-
Click File
-
Click Edit Story
-
Click Smart Insight Settings
- Click + Select Dimension Members
- Click GeoLocation
-
Click All Members
-
Click OK
- Click + Select Dimension Members
- Click Order ID
-
Click All Members
-
Click OK
- Click OK
🚩Lets view the smart insights panel again to see how the configuration has impacted the top contributor results. As we mentioned earlier the insights can be opened on any data point so we can select any point in the chat and launch Smart Insights to see the same results. This gives a story viewer the ability to run get smart insights on any data point.
-
Right Click on High Intensity
-
Select Smart Insights
🚩You can see there are no more redundant insights. Let's look at the location contributor.
- Expand the Germany Insight
- Scroll to the Bottom
🚩You can see now the top 10 location contributors. You like to add this chart to the page.
-
Click Copy
-
Click Copy to Page 1
-
Click Close
🚩The exploration of a value does not need to stop at the first data point, we can see that Germany is the top location contributor for High Intensity. However, if we want to drill deeper on Germany we can be running smart Insights on that sub set of the data. Again, this is particularly useful in view mode and for users without a good understanding of the underlying model.
ℹ️Let's copy the table we created in our previous story.
- Right Click on Files
- Click Open App in New tab to open MyFiles in a new window
🚩Find the file you saved
- Click ANA260_Your Initials
- Click Edit
-
Scroll to the bottom
-
Select the table
-
Click More icon
-
Click Copy
-
Click Copy
- Go back to our Current Story by changing the tab
🚩Press Ctrl + V to paste the table
ℹ️Quality Check! Does your story look like this?
ℹ️The range of insights provided in the smart insights’ panel are dynamic and vary based on the data point selected by the user. Earlier in this workshop we created a calculate of Gross Margin %. For a user viewing the resultant table it might not be clear how Gross Margin % is calculated. Smart Insights can help with this.
-
Rigth Click on the data point
-
Click Smart Insights
- Click to expand
ℹ️In this case as smart Gross Margin % is a calculated value we see different insights and the calculation is shown to the user.
- Hovering over a measure in the calculation shows the value.
🚩Smart Insights are available on the values contributing to the calculation, to allow the user to explore further and better understand the numbers. (Right Click on Gross Margin for smart insights)
🚩In addition to Smart Insights and Search to Insights, Smart Discovery provides automated data exploration. Let us try that now.
-
Click Tools
-
Click Smart Discovery
🚩If your user enters Smart Discovery for the first time this dialog may appear.
- Click Got it
ℹ️Welcome to Smart Discocery!
Smart Discovery offers you a viable way to use automated machine learning on top of your BI data, without losing precious analysis time on data preparation. Simply decide the business question you want to ask your data and let Smart Discovery analyze it for you, by running a machine learning algorithm. You can then explore the generated results to gain insights into your data.
ℹ️Smart Discovery helps you, as a business analyst, to understand the process that it’s using to analyze your data for you. It helps you to specify the right question, and quickly understand the generated results. You can refine the question by modifying the target or entity, filtering the dataset or excluding variables for the analysis. The Target is the measure or dimension you would like to know more about like Revenue or Customer Churn. The Entity defines the dimension or dimensions that describe the object in the data you would like to know more about, for instance customer or product. The entity describes the key that identifies each instance of that object. Smart Discovery will aggregate the data to the level described by the entity
🚩Our Business Question is to understand gross margin at the store level
- Click + Select a Dimension or Measure
- Click Gross Margin
- Click + Select a Dimension
-
Scroll to find Store
-
Select Store
🚩Before running the Smart Discovery, we can preview it and see what macro's we're going to be running the discovery on
- Click Preview
🚩The Smart Discovery preview provides an opportunity to confirm the Business Question before running the Smart Discovery. In this example Smart Discovery will explore Gross Margin at the store level. The dataset contains 453 stores and ~1.8M in gross Margin
- Click Run
It is important to note that the analysis is performed on a snapshot of the data at the time Smart Discovery is run and that the analysis is not updated automatically in response to updates to the data. All the content generated by Smart Discovery is dynamic and changes based on the underlying data.
ℹ️Smart Discovery Creates up to Four Pages:
- Overview: The overview page provides some high-level metrics and visualizations based on your business question
- Key Influencers: The Key Influencers page is generated based on the predictive model. The Key Influencers page lists (ranked from highest to lowest) up to 10 dimensions and measures that significantly impact the target.
- Unexpected Values: contains a table with the records that have been identified by the underlying machine learning model as unexpected. The table shows the actual and the predicted values along with the other corresponding dimensions.
- Simulation: It contains a listing of the key influencers and their corresponding values. You can simulate a row of data by selecting values for each key influencer. The target is updated based on your selections.
- Click Key Influencers
🚩To increase the users trust in the smart Discovery finding a clear Natural Language explaination is provided. This explains: • The users business question • The details of the data snapshot used • The automatic data preperation • The analysis quality • How to interact with the influencers
🚩Here we can see that the tool has identified 10 variables in the data that influence Gross Margin. Region has the greatest influence on Gross Margin. Scroll down to see more info.
- Scroll to the bottom
ℹ️As region has the strongest influence it is preselected. The average Gross Margin for stores by region and the distribution of stores by region are shown to provide an intuitive view of how region impacts gross margin at the store level. The input control allow the suer to switch between influencers.
- Scroll to the bottom
🚩How influencers combine to impact a target can also be interesting. The second set of charts shows the combination of the top two influencers. Using the two input controls you can combine different influencers to see their combined impact
🚩Let's look at the Unexpected Values
- Click Unexpected Values
ℹ️The unexpected values are stores in our data where the values for gross margin predicted by the predictive model differs significantly form the actual value in the data. These values can be interesting as these stores re different to the pattern detected in the overall dataset. This might indicate an issue with the data or store that have unusual behaviour.
Looks like we only have a few unexpected values. Let's move to the simulation page
- Click Simulation
- Click Got it
ℹ️The simulation page lets you adjust the input parameters and predict the impact of these changes. Let's try this out.
-
Scroll down
-
Expand Region
-
Click MEE
-
Click Discount field
-
Drag the slider as desired
-
Click Product Count
-
Drag the slider as desired
- Click Simulate
🚩See how your changes updated the impact of the influencers. You can also scroll up to see the overall impact on the Gross Margin
-
Click File
-
Click Save
-
Click Save
🚩Save the story in MyFiles
-
Type ANA260_Section 6_Your Initials
-
Click OK
You have completed the entire Augmented Analytics: Smart Assist Features section!
You are now able to:
- Use Search to Insight to find answers for your questions with natural language
- Use Smart Discovery to investigate your data and build a dashboard for your
- Use Smart Insights can help you explain more of your data