What is natural language augmented analytics

Augmented Analytics is the use of machine learning (ML) and natural language processing (NLP) to improve data analysis, data exchange and business intelligence. But what exactly can it do? Why is it better than traditional analysis? And how does it work? We answer these questions for you!

We start with the general explanation and the benefits of augmented analytics. If you're more attracted to the technicalities, feel free to jump over there!

In July, Gartner published the 2020 edition of the report, which now also includes "Augmented Analytics". This is far from surprising as it has been dubbed the "future of data analysis".

In the report, Gartner describes augmented analytics as:
"An approach that automates insights using machine learning and natural-language generation, marks the next wave of disruption in the data and analytics market."

Translated, this means that augmented analytics is an approach that automates insights with the help of machine learning and natural language generation and thus foresees the next wave of disruptions in the data and analytics market. While this is a great definition for data professionals, it isn't detailed enough for most people to understand what augmented analytics really means. So today let's clarify what “augmented analytics” really is - and why you should be interested in it.

What is Augmented Analytics?

Translated, augmented analytics means "advanced analytics" and to understand this concept we first need to identify the problem it solves. That means we have to internalize why generating insights from data remains a major challenge for almost all companies.

In the business world, everyone agrees that data analysis is good for businesses and has the potential to dramatically increase traction and revenue (if done correctly). The problem, however, is that data analysis isn't exactly the easiest thing to do.

Raw data

In fact, raw data in and of itself is completely useless to your company if you don't know how to handle it and how to interpret it. Let's say your online data shows that your earnings are down 10% from last month. But what does that really mean for you? Is this decline an industry trend? Is it because one of your advertising channels is not working? Or are there other reasons?

To get a better understanding here, you need to dig deeper into your web analytics, ecommerce, and social media data to find out what caused the decline in your earnings. Then you need to put those changes into a business context and identify those that you need to respond to immediately.

Knowledge analysis

If you return to the example of declining revenue, you may find that your social media ads are 10% less effective than the previous month and that you need to optimize your ad spend. That insight is now actionable because it is directly tied to an action that you can take to resolve your business problem. These actionable insights are extremely helpful as they serve as a guide to what priorities you should prioritize in your business.

However, to go from raw data to insights, you have to go through many technical steps, among other things
1) collect data from various sources,
2) clean data so it's ready for analysis,
3) carry out analyzes,
4) Generate insights and
5) Communicate these findings with the organization and translate them into action plans.

These steps are extremely complex and actually involve a lot of implementation effort in practice. To do this, I recommend that you hire data analysts to carry out these steps for your company.

First of all, data scientists and analysts are REALLY scarce and expensive to hire, making analytics extremely costly for smaller businesses. If this applies to you, you should take training courses or workshops to further educate your employees. Take a look at our offers or contact us via email.

Data scientist

This means that executives must work very closely with data scientists to ensure that the analysis results actually make "business sense". This robs managers of valuable time.

As everywhere, exceptions confirm the rule, even if only a very small part of the population represents them. Examples include industrial engineers with a focus on data science & analytics, as is the case with the Compamind department head.

In addition, the data scientists spend over 80% of their time in practice with simple mechanical things such as labeling and cleaning your data. This wastes the company's time and investment.

After all, data scientists are still human, so their attention spans and ability to do repetitive work are limited. Hence, a human can only analyze a small fraction (probably 10%) of your data that they believe has the greatest potential to provide you with great insights. That means you could be missing out on valuable insights in the remaining 90% - insights that could be critical to your business.

Almost all small and medium-sized businesses are still in the early stages of adopting analytics, despite a strong desire to leverage their data.

The challenge that companies face here is not going to go away anytime soon. We can't expect data scientists to be suddenly deposed by Martians to fix the talent shortage. Nor can we expect that analyzing this data will magically become easier for non-tech business professionals.

Augmented Analytics Engine

Augmented Analytics generates knowledge in a company through the use of advanced algorithms of machine learning and the automation of artificial intelligence. As a result, a company is no longer dependent on people.

An augmented analytics engine can automatically go through the data of a company, clean it up, analyze it and convert this knowledge into action steps for the executives or marketing experts. And all of this with little or no supervision by someone who is familiar with the subject.

Software comparison

This is the common misconception I hear from everyone outside (or even inside) the analytics industry. Software tools such as Tableau and SAS are tools that “support” the analysis. This means that these tools facilitate the analysis and communication of the results for the analytics in your company. However, they don't do the analysis for you, and they certainly don't eliminate the need for a business analyst or data scientist.

Augmented analytics, on the other hand, is designed so that analysis and business insights can be performed automatically with no (or little) oversight and can be used directly by marketers and business owners without the need for the help of a business analyst or data scientist. Because of this, this application is far more advanced and powerful than tools like SAS and Tableau.

Augmented Analytics with SAP (e-book)

More input on augmented analytics, predictive analytics, BW & Co.? Download our e-book now!

Augmented Analytics with SAP (e-book)

Tableau + SAS

On the technical side, tools like SAS and Tableau concentrate on offering extremely flexible interfaces. This way, analysts can easily perform any type of analysis on the platform and present the results beautifully. On the other hand, augmented analytics is much more focused on the ultimate goal of these analytical insights.

For example, while Tableau lets you create beautiful bar charts (without saying what the bar chart actually means for your business), an augmented analytics engine could simply tell you what your data says about your business. It may only display the relevant information that led the system to this conclusion.

Hence, augmented analytics software is much more focused on building a knowledge base of business information to identify business trends. And on using machine learning algorithms to identify these trends in a company's data.

Development of augmented analytics

In the past two years, Augmented Reality has grown VERY fast, so be excited about this data revolution! As a data scientist who works in this area, I am delighted that augmented analytics has made it into the business world.

I like to compare the development of data analysis technologies with the development of the automobile. For starters, they are both extremely complex systems that contain thousands, if not millions, of parts in order for them to function properly. However, despite the complexity of the technology, almost anyone in this country can drive a car. And despite the early critics who were convinced that this advance was impossible.

That's because most of the complexity is abstracted away by technology. Users only need to know the “data” that is relevant to them (e.g. how to use the steering wheel) in order to make decisions while driving. We are now making further progress to eliminate the need to “drive” so that users can simply take care of what is really important to them: getting from point A to point B as efficiently as possible. In the data world, we're nowhere near as advanced as the automotive industry's design and technology accomplishments (given that we've only been ten years into the data revolution, that doesn't surprise me).

Data sources

Data analysis software can integrate augmented analytics tools to process large amounts of data. Organizations can enter information from raw data sources into these platforms, which then clean, analyze, and return the key data for analysis. The use of machine learning and NLP gives augmented analytics tools the ability to understand and interact with data organically and to identify valuable or unusual trends.

The field of data analysis is complex and generally requires a data scientist who can extract any value from large data. Part of this complexity is due to the need to collect data from a number of different sources such as web analytics, marketing communications, and social media posts. Gathering the data is only the first step, it must also be prepared for analysis by organizing and refining it before the analyst or data scientist can gain useful insights. The results must then be communicated to the organization along with action plans in order to use these findings.

Data preparation

Due to the manual effort required for these tasks, data scientists are currently in great demand and can be impractical & expensive for some companies. It is estimated that a data scientist spends up to 80% of their time collecting, preparing and cleaning up data. At this point, augmented analytics can be used.

With the addition of machine learning to data analysis, many of the time-consuming tasks of data collection and preparation can be performed quickly, automatically, and with fewer errors. This allows data scientists to spend more time looking for usable insights. The real, ultimate goal of augmented analytics, however, is the complete replacement of data science teams with AI. The entire analysis process from data acquisition to business recommendations is taken over by the decision-makers.

To be very clear, you could ask the augmented analytics tool to find reviews online about one of your products and tell you what you should improve in order to sell more of it. The machine responds with a clear textual answer and some convincing diagrams.

Augmented Analytics Services

Augmented data preparation

Augmented data preparation enables business users to access meaningful data to test theories and hypotheses without the help of data scientists or IT staff. It enables users to access important data and information and allows them to connect to various data sources (personal, external, cloud and IT-managed).

Users can summarize and integrate data in a single, unified, interactive view and automatically clean, collapse and clarify suggested relationships, JOINs, type casts, hierarchies and data so that they are easier to use and interpret. Integrated statistical algorithms such as binning, clustering and regression are used to suppress noise and identify trends and patterns. The ideal solution should strike a balance between agility and data governance to ensure data quality and clear watermarks to identify the data source.

Augmented data discovery

Augmented Data Discovery is an emerging BI functionality for the automatic preparation and organization of company data for self-service BI. This is a particular challenge for unstructured data from sources such as email, social media channels, IoT feeds, and customer service interactions.

Traditional BI tools have supported basic capabilities for merging, manipulating, and transforming structured data. The advanced data discovery can build on these basic functions with advanced data preparation and automated pattern recognition for self-service BI, according to the research company Gartner Inc. The advanced data preparation streamlines processes for data profiling, quality management, data cleansing, modeling, enrichment and labeling of metadata in a way that supports reuse and management. Automatic pattern recognition builds on traditional BI tools to support complex, large data sets.

Conversational Analytics

Conversational Analytics is a technology that transcribes speech and converts it into data. The data is then structured in such a way that the conversations can be analyzed for insights. Conversational analysis solutions usually consist of a transcription engine that converts speech into data; an indexing layer that makes the data searchable; a query and search user interface that enables the user to define requirements and perform searches; Reporting applications to present the analyzes, often in graphical form; and business applications provided by vendors to support users with specific needs.

Automate data analysis

Augmented Analytics makes this easier by automating the process of analyzing data and generating insights.
It identifies trends and explains what they mean in practical terms for a company through clear visualizations and neatly packaged trends. A characteristic of augmented analytics that differs from other technologies is the ability to perform “natural language” generation that unwraps complex jargon and provides insights in simple, understandable terms like “56% of leads were generated from PPC ads”.

In addition, augmented analytics eliminates the limitations that human bias can bring. It's not tied to a specific research question, and it gives companies the freedom to uncover the many layers of insights their data has to offer - even insights that weren't considered in the first place.
All in all, this means that executives can focus their attention on the strategic side of things instead of being locked into routine arithmetic tasks.

Some estimate that once augmented analytics is at its peak, it will eliminate the need for data scientists and analysts. However, experts agree that in reality the most likely outcome is that these roles will simply evolve and focus more on specialized problems and embedding models in enterprise applications, working hand-in-hand with augmented analytics to perform their roles more efficiently.

Business processes

The purpose of augmented analytics is not to replace the decision-making process, but to support it.
The beauty of augmented analytics is that it does not replace the need for certain technical roles within an organization, but that it actually helps harness human expertise.

Instead of spending months preparing and analyzing data, leaders get the keys to unlocking clear trends, visualizations, and insights at will. This helps in supporting business decisions and enables the experts to do their work with greater accuracy.

With augmented analytics, human expertise is actually becoming more crucial than ever. The abundance of knowledge creates the temptation to get caught up in a "Shiny Object Syndrome" as each knowledge has potentially equally exciting opportunities for exploration. The experts need to combine their expertise with the initiative to sort these nuggets and filter out only those that align with the overall business strategy.

These broader insights will also prompt the experts to dig deeper to create even greater value for the business, and they will be instrumental in ensuring that their organization is data-driven.

Ultimately, augmented analytics will remove the boring, robotic processes associated with BI and enable employees to focus on being human.

Augmented Analytics with SAP (e-book)

More input on augmented analytics, predictive analytics, BW & Co.? Download our e-book now!

Augmented Analytics with SAP (e-book)

Conclusion

Here are a few reasons why you should consider advanced analytics and data preparation for your business:

  • These solutions enable data scientists and the IT community to focus on strategic issues and specific projects
  • Accessible Augmented Analytics creates Citizen Data Scientists and improves accountability and empowerment
  • Advances in intelligent data discovery and other sophisticated techniques and solutions can have a positive impact on ROI and TCO
  • These solutions result in better decisions, more accurate business predictions, and measurable analysis of product and service offerings, pricing, finances, manufacturing, and other aspects of the business
  • The advanced data preparation and associated tools improve user adoption, data popularity, social BI integration, and data literacy

But, to quote Gartner, Augmented Reality 2017 was the future of data analytics because it brings us closer than ever to the vision of "democratized analytics" because it is cheaper, easier and better. In 2020 we will finally see that more and more companies of all sizes are using and benefiting from analytics. Two years after this study, we can now say that augmented reality has arrived in the here and now and that we can help you introduce it.

Free web session augmented analytics

Would you like to finally make data-driven decisions? Let's talk about your challenges and wishes in a free web session!

Free web session augmented analytics

Please leave us your name and contact details. Then we will contact you.

Swen Deobald

My name is Swen Deobald and I am an enthusiastic SAP Analytics consultant. As Compamind's head of department, my team and I support you in all matters relating to SAP Analytics, Business Warehouse, BusinessObjects and the SAP Analytics Cloud.

Do you like this article? Subscribe to our news and blog posts!

Enter your e-mail and you will regularly receive a compilation of our current blog posts.


You might also be interested in:


You might also be interested in

Time Series Forecasting enables reliable decisions to be made using time series and forecasting methods. Data from time series contain a value over time, e.g. sales by month or call volume by week.

SAP Analytics Cloud (SAC) has been around for some time and is primarily a tool for cloud-based data analysis, especially for customers who want a seamless connection to their existing cloud-based solutions. From version 2019.8 the [...]