Jim, the Sales Head of one of the leading pharmaceutical firms receives the monthly sales dashboard from his Business Intelligence team. He has to explain to the board why the sale of their top-selling drug has been at an all-time low for the past two months in the United States. As if this task was not challenging enough, Jim must also present the reasons and strategies for rebooting the sales in next 12 hours to the board.
Now, let’s look at how Jim would approach the same business problem at hand in different eras.
2012: Canned Reporting Era
Jim formulates a few hypotheses that could be the reason for the drop-in sales and asks the data scientists in his organization to test those for him. While waiting for the results, Jim spends the rest of his time connecting the findings that need to be presented at the board meeting.
2016: Self Service Era
Jim has a ready-made dashboard which updates in real-time. He tests the hypotheses formulated by him and stops when few of his hypotheses are confirmed. This results in overlooking other important factors and a suboptimal conclusion. However, Jim has less dependency on data scientists and more time to connect the findings to construct a coherent story.
2020: Augmented Analytics Era
Jim asks a query, “Why are sales at an all-time low in the United States for the past two months?”. The platform automatically analyses the data from multiple sources and presents the insights along with actions that need to be taken to prevent the loss of sales, thus eliminating the possibility for any human bias to get the most accurate results. Jim then formulates the steps that need to be taken for reviving the sales of his company.
This is how augmented analytics-driven platforms are redefining the way we make data-driven decisions. A business person without any knowledge of data mining or statistics can derive the true value of data in real-time with minimum or no effort.
What is Augmented Analytics?
Augmented Analytics (mentioned as “Augmented Data Discovery”) featured in Gartner’s “Hype Cycle”, published in July 2017, for the first time which they claimed to be the “future of data analytics”.
Gartner’s report entitled “Augmented Analytics Is the Future of Data and Analytics”, Published 27 July 2017, by Rita L. Sallam, Cindi Howson, and Carlie J. Idoine, says Augmented Analytics will mark the next wave of disruption in the data and analytics market.
In the report, Gartner defines Augmented Analytics as:
“an approach that automates insights using machine learning and natural-language generation”
- Augmented data preparation: Uses ML to find anomalies in data, check for data quality, standardizing and profiling data
- Augmented data discovery: Uses ML to automatically find, visualize and narrate relevant findings without having to write any algorithm or build any model
- Augmented data science: Automates key aspects of advanced analytics such as feature selection.
Current Data Analytics vs Augmented Analytics
|Analytics Workflow||Current State||Future (Augmented Analytics) State|
|Prepare data||Manual data preparation, data quality, enrichment cataloging||Algorithms detect schemas, profile catalog and recommend enrichment, data lineage and metadata|
|Find patterns in data||Manual exploration of data using interactive visualization’
Manual feature engineering and model building
Algorithms find all relevant patterns in data
Models are autogenerated
|Share and operationalize findings||Dashboards, storytelling, collaboration
Depend on user to interpret results
|Insights are narrated in natural language or visualizations to focus user on what is more important and actionable
Can be embedded in apps or conversational UI
Source: Gartner report on “Augmented Analytics Is the Future of Data and Analytics”
How will Augmented Analytics redefine BI and Analytics market?
Analytics market which is the core of digital business has evolved by leaps and bounds over the last decade. Most of the organizations in the world, if not all, are dependent on data to uncover insights and make data-driven decisions. Storing and mining large volume of data has become easier, tools across analytics landscape have become more agile and easier to use, however organizations are still largely dependent on data scientists to prepare data for analysis, analyse data, build statistical models on the data and uncover hidden meaning out of it.
BI and Analytics consumption has evolved from canned dashboards, where there was a time lag between the business requirement and delivery to a self-service era. Now business people and analysts can play with the tool and analyse data on their own. However, due to a large volume of complex and cross-functional data, it becomes almost impossible to check for all possibilities and arrive at the most relevant findings. The analysts in the process are prone to confirmation bias which leads to a suboptimal conclusion which can adversely affect important data-driven strategic decisions of organizations.
The answer to all problems in the current BI and Analytics landscape lie in augmented analytics-driven BI and Analytics workflow. Here are a few ways on how augmented analytics can help an enterprise make faster and more accurate data-driven decisions:
- Lesser dependency on data scientists
McKinsey’s Report entitled “The age of analytics: Competing in a data driven world”, published December 2016, predicts that there would be a shortage of 250,000 data scientists by 2024 and retention will continue to be an issue faced by organizations. Augmented analytics would help reduce dependency on data scientists and create Citizen Data Scientists which would increase accountability and empowerment.
“By 2020, due largely to the automation of data science tasks, citizen data scientists will surpass data scientists in terms of the amount of advanced analysis they produce and the value derived from it.”
- Focus on strategic issues and developing an action plan
Data scientists spend almost 80% of their time on data preparation and cleaning data. (Read here). With the help of machine learning driven analytics workflow, data preparation and data discovery will be automated giving more time to business experts for focussing on core strategic decisions.
Source: CIO Insight
- More accurate and actionable insights
Data scientists are data experts at the end of the day and not business experts. One of the biggest barriers to derive full value of data is combining data knowledge with industry and functional expertise. A visualization tool is an effective way to communicate patterns in data, however, most of the findings require further deep dive and statistical analysis to determine whether they are accurate and actionable. Augmented analytics would remove human biases, automate statistical tests and give insights by combining cross-functional data sources that would lead to more accurate and actionable insights.
- Bring in a new dimension to conversational analytics
A new paradigm in BI and Analytics is conversational analytics where business users can raise query such as “Alexa, which state had the highest sales in the last month?”. Imagine a state where the business user doesn’t even need to raise a query. The augmented data discovery platform analyses the data and finds significant insights which otherwise the business user could have missed. Combining NLG with the platform automatically presents a written or spoken context-based narrative of findings in the data, alongside the visualization, that inform the user on the most important areas to act upon.
Source: KO Marketing
Augmented analytics is not mature yet but is bound to grow at a very fast rate in the next couple of years to disrupt the BI and Analytics market. The demand of skilled data scientists is ever increasing. Augmented analytics is an answer to this shortage of skilled data scientists that can iteratively perform the data-to-insight-to-action activities. Organizations need to adopt this as the platform is mature to stay abreast and relevant in the industry.