Cognos Analytics

 View Only

BI Modernization with Cognos Analytics

By NORBERT Bracke posted Mon September 23, 2019 08:04 AM


If you are legacy Cognos user, you should become familiar with the new capabilities that have been introduced with Cognos Analytics. Modernization is not a new term but let’s review it in terms of how Cognos Analytics can help drive that initiative. Typically, Cognos has been used for operational style reports and some adhoc reporting. Worst use case Cognos has been used as an ETL delivery mechanism; sound familiar? With Cognos Analytics releases features have been and continue to be enhanced that provide functionality to support a more modern analytics environment. Even with the new capabilities the Legacy Studios still exist for now which will make the transition to Cognos Analytics easy. Investments in Cognos Analytics are improving the ability to provide Descriptive Analytics - What Happened, Diagnostic Analytics - Why did that happen and Predictive Analytics - What will happen.


Let’s start with a review of the Cognos Analytics Report. Although the Report authoring environment look and feel has changed there still exists a Compatible Viewer which is using the same code stream as Cognos v10.2.2 so that all reports that have been upgraded to Cognos Analytics will render the same.
What has Picture2.pngchanged is the introduction of the Interactive Viewer which is enabled by default on any new report but can be controlled in the properties of the report. It is not recommended to mass change all your reports after a migration from Cognos 10.x so that each report can be carefully considered for capability with the interactive viewer. Things to consider are slow running reports, reports that have javascript and the visualizations used. Previously javascript was implemented using an HTML item, that has changed now to use files stored on the server. This new approach improves reuse and better access to the report data. Read this for general overview of the new approach on Javascript in Reports here. The new interactive viewer presents an on-demand menu to perform actions on the report, these interactions are similar to what was available inPicture3.png some of the Legacy Studios but presented in a new UI. These actions are valuable for an interactive experience when exploring the data but if the data access is slow then this would be frustrating hence, we need to consider the performance before enabling the interactive viewer. Additionally, new visualizations in 11.1 include new chart types and browser side data for improved interaction and performance. This local data allows for filtering in the browser without a return request to the database. A list of these features on report interactions can be found in the documentation here and the new 11.1 charts here.


The behavior that modernization is attempting to establish is self-service Data exploration. Certainly, a report with the interactive capabilities enabled will allow a user to filter, slice, sort and perform some analysis on existing data defined in a report. Perhaps a better tool for data exploring is the Cognos Analytics Dashboard. A core capability of the Cognos Analytics dashboard is visually exploring your data. This highly interactive experience allows the user to drag data elements onto the canvas and visually see the data in real time. How the data is displayed can be controlled by the user or automatically by the AI capabilities of Cognos Analytics.Picture5.png
Picture4.pngSome of the additional smarts built into the visualizations include automated Insights and Forecast.  Also, part of the AI component delivered with Cognos Analytics is the Assistant where you can use Natural Language to interact with your data and the dashboard. 

More information on Dashboards can be found here and the on Assistant can be found here.


Data Journalism is another component of modernization in this new digital era where complex data is explained using compelling visualizations.  In Cognos Analytics this can be implemented using Cognos Analytics Stories. With Stories you can create animated presentations using live data connections based on the insight collected from other components of Cognos Analytics. More information on stories can be found here.


To accelerate the modernization efforts, we also need to include Data Science capabilities. Data Science is a trending topic yet can be complicated to endeavor. Cognos Analytics includes two capabilities to bridge the gap of traditional BI to modern analytics exposing AI and Data Science.


Cognos Analytics Exploration is a simple interface to find patterns and relationships in your data using Picture7.pngPicture6.pngstatistics. A user can start with data, a Dashboard or a Story. Using AI additional details on the data and related visualizations are recommended for deeper analysis. Also, the Assistant is available which presents a natural language interface to dialog with the AI engine.  More information on Exploration can be found here.


Picture8.pngCognos Analytics Notebooks are also available for users with more experience in Data Science or someone who simply wants to begin a data science journey.  Currently the Jupyter Notebook in Cognos Analytics support Python scripts and can read and write data using a data source defined in Cognos Analytics. The output of a Notebook cell can be embedded in a Cognos Analytics Dashboard, Story or Report.


In summary the goal of analytics is to enable data exploration to make better business decisions. How that is accomplished can be done leveraging a variety of tools. The bigger challenge will be data clarity but if the tools used allow fast access to data then the journey has begun. With AI embedded in a tool then we can also begin an intelligent assisted evaluation of the data which may lead to results not easily obtained by random adhoc exploration. The evolution of Cognos Analytics is exciting and may bridge the gap between a “mathy” data scientist and a business user who is just trying to understand the data to make decisions. Perhaps the business user can find something interesting and then leverage the experienced data scientist to dig farther and find an explanation of the finding.


Here are some YouTube videos on many topics around Cognos Analytics.