The term Business Intelligence was first used in 1865 and was later adapted by Howard Dresner in 1989 to describe business decisions based on data analysis. In the 1990s and thanks to the evolution of data warehouse technologies, data mining was born for discovering patterns within large data sets. Analyzing data in non-traditional ways soon provided results that were not only surprising but beneficial and businesses began to predict customer needs based on looking into consumption patterns. Nowadays, data analytics form the digital core of leading data-driven organizations, either in the world-renowned Fortune 500 or in SMBs.
Due to the vast number of data sources available and the marketers’ willingness to understand how to get the most out of them, data volumes are exponentially increasing, together with the complexity of taking effective strategic decisions.
On the other hand, preparing data for its analysis has always been a time-consuming activity. According to a Gartner report, data scientists spend up to 80% of their time performing mechanical and/or repetitive tasks such as labeling and cleaning data, as opposed to making sense of it. Moreover, “subjectivity” has also been inherent throughout exploratory phases in traditional business intelligence and analysis processes, in which humans traditionally played a spectator role. All this has brought about a scenario where report consumers receive biased hypotheses, neglecting potential key findings and drawing incomplete or incorrect conclusions. Besides, there is a strong dependence on data scientists throughout the process and they may not be precisely experts on specific business fields, such as Marketing, and might not find the most useful insights or come to meaningful conclusions. Any organization in the race to achieve full digital marketing maturity will rely on an automated data-driven framework to make decisions continuously.
In this article, we will start to unveil what Augmented Analytics is. The rapidly-growing data analysis practice was born as a response to the aforementioned scenario and it is a key starting point for marketing teams that strive to achieve multi-moment maturity. That is, the ability to deliver the relevant content for the target consumers at multiple moments across the end-to-end purchase journey. Once they manage to do so, cost savings of up to 30% and revenue increases of as much as 20% have been reported in a recent research lead by The Boston Consulting Group where more than 200 global brands from 10 industries were studied.
Augmented Analytics incorporates Data Preparation, Automated Insights Discovery, and Natural Language Processing, all accompanied and surrounded by Machine Learning (ML). Augmented Analytics is a term primarily coined by Gartner, sometimes referred to as Augmented Intelligence by some vendors, but in a nutshell, it is the modern and natural evolution of the traditional analysis process. Let’s take a look at the pillars that constitute this practice, and that not only benefits the Marketing team but any other organizational business unit.
Automated Data Preparation empowers anyone with access to meaningful data to test theories without the assistance of data scientists or IT staff. It allows users to connect various data sources, whether these are personal, external, Cloud or IT-provisioned and integrate this data in a uniformed view.
Insights Discovery, Visual-based Data Discovery or Smart Data Discovery enables users to leverage sophisticated analytical techniques without specialized help, allowing employees to find relationships, identify trends and patterns, receive suggestions for data visualization and forecast results.
Natural Language Processing (NLP) is the well-known sub-field of Artificial Intelligence that in this particular case assist the understanding and comprehension of large datasets by providing a semantic understanding or written summary of the insights and data of interest. In the background of an analysis or reporting platform, the NLP engine will be performing computations on large blocks of data without any manual effort. Insights, therefore, become accessible, both easy to understand and to share.
What’s the end result then? The updated process completely transforms how content is developed, consumed by data specialists and shared within an organization. At the start of the chain, during the exploration phase, businesses are able to decrease the time spent on tools with robust data-drilling functionalities as both the data preparation and discovery processes are accelerated. In addition, personal opinion is taken out of the whole analysis stage, allowing users to act on data more efficiently, which enables the exploration of fact-based hypotheses and hidden patterns. Ultimately, the traditional data analysis process is democratized and accessible to every employee since end-users not only have easy access to objective discoveries but also to the insights distribution phase. The following diagram visually describes this scenario where ML, AI, Prediction and Automated Insights work in conjunction, resulting in business instructions for senior roles, executives, management and salespeople.
In a recent report, Gartner highlights how Augmented Analytics is not a futuristic concept yet a current strong market disruptor and a technological trend, it represents a third major wave for platform capabilities in this space, and it has been considered one of the Top 10 strategic technology trends for 2019. CDOs and CIOs are adopting this growing practice as part of their digital transformation strategy as it enables the delivery of advanced insights to a broader population while embedding a virtually-automated data scientist role within a reporting scenario.
The following image describes how the traditional, self-service processes evolve into automation.
Sweetspot takes pride in assisting organizations to leverage their data assets, and this is why we totally dedicate our work to empower individuals with access to data to test their own theories without a data scientist role in between. Augmented Analytics is firmly embedded in our platform, allowing business users to consume auto-generated visualizations and narrate relevant findings using collaborative interfaces, such as Sweetspot Storytelling elements. Even though we have provided forecasting capabilities in the past supported by the ARIMA algorithm, we are now taking a step further by providing automated actionable predictive insights. These insights are contextualized natural-language-generated narrations (voice or text) and guided recommendations, that if combined with our mobile apps, it enables users to even act on them using portable devices such as a mobile phone or a tablet. To sum up, our users can confidently perform analyses and generate business insights automatically, not relying on the supporting role of a business analyst or a data scientist. It all can be done on the go.
In two years’ time or even less, Augmented Analytics will be a must-have for any firm that aims to act on their data, and from Sweetspot, we have eagerly encouraged organizations to conduct a pilot project to assess the feasibility of adopting this practice. The first stage of implementation should always be a re-cap of the existing manual, time-consuming data analysis tasks that could be polluted by biased decisions. Secondly, we recommend running the new processes in parallel with the existing ones in order to compare conclusive insights, differences of inaccuracies or outcomes each process presents. This assessment will allow companies to objectively measure improvements against the traditional scenario.
The ultimate goal is to make analytics part of our day-to-day and have everyone consume data without really realizing it. Few years down the road, end users most likely won’t work with business applications that are not enabled in some way by Machine Learning; analytics is no different.
What about you and your team? Have you started to take a look at how you can benefit from Augmented Analytics?
Not Another Dashboard.