Skip to Content
Technical Articles
Author's profile photo Milen Ann Abraham

Smart Insights and Smart Discovery in SAP Analytics Cloud


Smart Insights and Smart Discovery are two machine learning features in SAP Analytics Cloud. With the power of machine learning, they help users to take advantage of advanced contribution, classification and regression techniques. These features help anyone to find out hidden patterns and complex relationships within their information, even without any data science knowledge or experience. These are powerful machine learning capabilities that helps business to make quick decisions with SAP Analytics Cloud.

Smart Insights

What is smart insights ?

Smart Insights pick ups a data point, variance in data and examine what is behind that data. It helps to find quickly what is behind a particular item. It can add contexts to your visualization which helps in understanding what is going on.

Smart Insights finds out the top contributors of a selected value or variance point. Top Contributors are the dimension members that provide the highest contribution to the data point being analyzed.

Why Smart Insights?

The benefit of Smart Insights is that, it helps the business users a major time saver when looking for quick answers to a particular value. So without the use of Smart Insights, a business user would have to manually pivot the data to identify the members from each dimension that contribute most to the data point.

How Smart Insights work ?

When a particular data point is selected, machine learning calculations run on information that is of the same nature as the selected data point. For example, if the selected data point is Total Revenue, the top contributors are based on Total Revenue. It analyzes the dimension in your selected data and looks for members in these dimensions that influence the selected value.

To run the Smart Insights, choose a data point on a chart to display the quick-action menu and select the light bulb symbol.

Select a data point from a chart -> Quick Action Menu ->

Example :

In this scenario, Smart Insights are used to explain the top contributor to the Total Revenue of Sales for a particular organization.

  • In the created story go to ‘More Actions Button’ -> ‘Add Smart Insights’

  • Now by running Smart Insights we quickly see that Central region is the top contributor to our sales.

  • Even by using the light bulb symbol on top right of the page you can search Insights, For example you can search for the top 2 regions which contributed more to the Total Sales.


  • It will display the top 2 contributors for Total Sales as shown below :

Smart Discovery

What is smart discovery ?

Smart Discovery is a very powerful feature of SAP Analytics Cloud that uses machine learning to analyse and explore your data and uncover valuable insights. SAP Analytics Cloud’s smart data discovery feature, helps in saving time by running automated machine learning algorithms in the back end to find out correlations between your dataset elements against the target metric, for example KPIs like revenue, time to fill days, sales, etc. When you make a few clicks, you can not only get all the key influencers for your target but you can also see the impact of other variables and view the data anomalies and run what-if scenarios, analyse the patterns in data and use historical data to predict future outcomes.

Why Smart Discovery?

Smart Discovery in SAP Analytics Cloud helps business users to interact with insights, in form of intuitive charts and natural language processing (NLP), to make faster and better decisions and share the new found valuable insights with their organization.

How Smart Discovery works ?

The user selects a measure or a dimension.

If a measure is selected, a regression model is built and if a dimension is selected, a classification model is built. If a dimension is selected, then Smart Discovery focuses on the members of classification (target) group selected by the user.

When the user selects Run, Smart Discovery will begin to build and test multiple test models using automated machine learning technologies. Smart Discovery will then select the best model based on accuracy, robustness and simplicity. This model will be used to generate the 4 pages of the story – Overview Page, Key Influencer Page, Unexpected Values Page and Simulation Page.

  • Overview Page – This page provides visualizations to summarize the results for the target dimension or measure.
  • Key Influencer page – Key influencers are measures and dimensions that influencethe results or outcomes. They are recognized from the information in your selected model using classification and regression techniques. Classification techniques are used to identify dimensions that segregate results into different groups of results. Regression techniques identify relationships between data points in order to predict future results.

Key Influencer page shows variables that are inter related to, that have the most influence, on the target. This page lists (ranked from highest to lowest) up to almost ten dimensions and measures that significantly impact the target of the discovery.

  • Unexpected Values Page – The Unexpected Values page shows the information about outliers. This page is displayed only if there is unexpected values. The table shows records where the actual amount differs greatly from what the predictive model (expected values) would expect. The scatter-plot displays these outliers to compare expected values versus actual values. The bar chart compares the expected values and actual values for selected record.
  • Simulation Page – When the Smart Discovery target is a measure, a simulation page is generated. It allows us to test hypothetical scenarios. The page uses the key influencers in an interactive what-if simulation. To the right of the page, a listing of the key influencers and their corresponding values is displayed. User can modify a value and simulate its impact. Choose the value and use either the displayed slider or radio buttons in the page to specify a new value. Each time a value is changed, a number flashes to show the percentage change from previous set of values. The chart displays the contributions of each of the key influencers based on the chosen values.

After you have finished analyzing the results from the Smart Discovery, you can:

  • Save the smart discovery as a part of the new or existing story.
  • Share this story with the other users in your organization.

Example : Analyzing Contractor Data to meet Budget

In this example, we use a scenario where an organization constantly hires contract workers. We would analyse the factors that have a positive and negative impact on contractor completing their assignment with their given budget. With the use of Smart Discovery, the organization can find out the influencing factors and then can take adequate actions to ensure contractors finish their work within their budget.

  • Create a “New Story” and import your data
  • Create a “New Smart Discovery”



In this example, the ’Percent_of_Budget_Consumed’ column in the data shows how much of a contractor’s given budget was used for their project. We shall find out what data or factors are influencing the ’Percent_of_Budget_Consumed’ field. Smart Discovery’s interface allows us to easily exclude any measures and dimensions that are not required from the analysis.

  • Select ’Percent_of_Budget_Consumed’as the measure
  • Click “Run”button

When Smart Discovery runs, predictive technologies run behind, going over different algorithmic models to help us find the one that best fits our data.

Key Influencers :

The bar chart in Key Influencer page shows the top influencers of the ’Percent_of_Budget_Consumed’. By clicking them, we can gain further insights about specific influencers’ impact. It shows up ’Supervisors’ that have the highest influence on whether or not contractors finish with their ’Budget’. On an average, Eric and Yuru’s contractors collectively exceed the budget and so the organization can take up needed actions.


Unexpected Values :

Next, we can analyze the unexpected values in our dataset to find out the differences between predicted (expected) and existing (actual) values of the Budget_Consumed. The Unexpected Values page gives the below information :

Simulation Analysis :

The simulation feature can be used, to see how much we can expect in Percent_of_Budget_Consumed. Based up on the selected criteria, we can see the contribution by Influencer on the possible Percent_of_Budget_Consumed.Shown below, under the selected Supervisor along with all the remaining fields, we can see that there is an expected budget consumption of 93.89%.

Sharing Insights :

Lastly, after gaining and gathering all these new insights through our Smart Data Discovery analysis, we can share this information with our colleagues. We can simply choose any of the charts we need and pin them to any of pages for the purpose of sharing.

  • Click on any chart you want to share
  • Click “Copy to Page”
  • Choose the desired page to pin the chart for sharing


Smart Insights and Smart Discovery features of SAP Analytics Cloud thus allows users even without any data science background to gain insights and analyze the hidden and complex relationships and patterns in the information which helps them in taking better decisions.


The images are referenced from :

Assigned Tags

      1 Comment
      You must be Logged on to comment or reply to a post.
      Author's profile photo Kun Peng
      Kun Peng

      Thanks, very interesting blog 🙂 I would like to know where could we find more information about the algorithm used behind Smart Discovery Key Influencer analysis. Not the general document for regression or classification model but more specifically for Smart Discovery regression and classification.