By Dr David Iacuone, Manager, Data Science & Business Intelligence, Health Metrics

Exploring the Smart Analysis and Signal Features of the Yellowfin Business Intelligence Suite – Version 8.0+


At Health Metrics, the Yellowfin Business Intelligence (YF) suite is tightly coupled to our flagship product: eCase. Because eCase is a real-time, on-line system, the new features are a hand-in-glove fit for it. YF is a powerful suite of tools that are utilised by our clients to measure performance and track against a number of metrics and/or dimensions (Eg. Prescribed Budgets).

This blog will explore the intrinsic benefits of using certain features of YF. Specifically, the Assisted Insights and Signals feature. Please note that this blog provides a brief overview of implementing these features and also looks at their benefits to the end-user. For a more comprehensive tutorial or simply more information on the matters contained in this blog, please contact Health Metrics on


Smart Analysis

In the past, data analysts would spend most of their time manually creating reports and responding to a variety of requests from stakeholders. They would follow many long-winded steps for manual data discovery. The goal of Smart Analysis is to bypass this arduous and often tedious process and obtain the same types of visualisations and insights that would take an analyst protracted amounts of time to generate, with just with a few simple mouse clicks.[1]

The Smart Analysis feature is implemented at the draft stage prior to publication of the report. You can either ‘explain’ or ‘compare’ variables. Firstly, we will review the ‘explain’ function.



The ‘explain’ feature helps you to understand the data which drives changes in your content[2]. Once you have created a table, simply click on the orange circle with the lightning bolt slash (highlighted below) and you will be taken to the ‘Insights Wizard.’ Select the ‘explain metric’ option and choose the variable you want to analyse.

Smart Analysis Feature

Figure 1: Initiating the Smart Analysis Feature


The Insights Wizard presents you with a range of data visualisations, graphs and written analysis. I have been working in the market research industry for many years and the range of analysis generated by the product is similar to what you would normally consider when formulating an analysis for a report. This feature allows you to generate a range of analysis and then pick and modify the parts you like the most or choose the parts that best support the business case you are trying to formulate. The types of analysis generated from this auto feature include:

  • Variable correlation;
  • Standard deviation;
  • Averages;
  • Totals;
  • Highest value;
  • Lowest value;
  • Top three performers; and
  • Bottom three performers.

An example of the results of the ‘explain’ feature are presented below (highlighted in red).

Figure 2: Results of the Explain Analysis



In the Analysis Wizard, the user also has the option to select the ‘compare’ function. This is similar to the ‘explain’ feature but, as the same suggests, focuses on juxtaposing variables and shows you how they impact on one another. With the ‘compare’ feature, you can perform the following functions:

  • Comparison of variables,
  • Comparison of dates; and
  • Comparison of dimensions (ie: categorical & descriptive information).

In order to setup a comparative analysis, you go through the same steps as the ‘explain’ function, but upon choosing the metric that you want to investigate, you select one of the three ‘compare’ options instead of ‘explain’. Then you select the two variables that you want to compare. In the example below, I am looking at a group of aged care facilities and comparing the ‘available number of beds’ versus the ‘number of occupied beds’. An example of the results of the ‘compare’ feature are presented below (highlighted in red).

Figure 3: Results of the Compare Analysis

Figure 3: Results of the Compare Analysis

This comparative analysis provides us with the following types of data:

  • Variable comparison;
  • Standard deviation;
  • Totals;
  • Top three performers;
  • Bottom three performers;
  • Highest value; and
  • Lowest value.



Signals is a powerful feature of the Yellowfin software package. It is a form of system monitoring where Yellowfin will perform real-time observations of your data and show you any significant changes over a specific period of time which could be days, weeks, months or even years. For example, if we monitor the data for a hypothetical aged care facility over a six-month period, we might observe that three residents have a sudden spike in their blood-sugar levels. The Signals feature will notice this change, treat these three residents as statistical outliers, diagrammatically represent this change and notify the end-user of these fluctuations. As a general overview, the Signals feature is designed to monitor the following types of changes:


Table 1: Types of Signal in Yellowfin [3]


The Signals feature is setup during the view creation stage. You initiate the signal by clicking on the little robot head at the top of the screen (shown below).

Figure 4: Initiating a Signal

Figure 4: Initiating a Signal


Select the signals tab and then hit the ‘create new signal’ button. Then you can select the type of signal you want (outliers, period comparisons or trend changes). From there, follow the screen prompts and your signal is ready. Do not forget to save your report before exiting. Once you have created the signal, it will appear in the main menu under the ‘signals’ option as shown below.


Figure 5: Locating Completed Signals

Figure 5: Locating Completed Signals


Once you open the signal, you will see a data visualisation similar to the one below. The legend underneath the graph explains what the different colours in the bar chart represent.

 Figure 6: Visual Display of Signals

Figure 6: Visual Display of Signals

If you click on the ‘analysis’ tab, you can see a range of graphs which highlight different aspects of your signal. This is where the signals feature really shines and you can discover some significant insights at this stage of the process.


Figure 7: The Auto Analysis Function Embedded into Signals

Figure 7: The Auto Analysis Function Embedded into Signals



The graph above shows incidents involving acts of physical aggression in the aged care facilities of an undisclosed provider. The two time periods covered are:


05/04/19 to 12/04/19; and

05/05/19 to 12/05/19.


This graph shows that for a specific aged care provider, acts of violence which occurred in lounge rooms had a relative increase of 100% from one period to the other.


Final Thoughts…

Market research and CRM data analysis have come a long way in the past 20 years. Analysis has moved away from the human agent and ideas such as machine-assisted automation and sophisticated algorithms are being incorporated into data analytics. Business intelligence tools can assist organisations in standardising reporting practises, minimising labour and reducing human error typically associated with report preparation. The Smart Analysis and Yellowfin Signals are powerful tools which aid organisations in these areas.





[2] Yellowfin (2017) Yellowfin Release Notes ver 7.4. Retrieved from


Share article