Six Characteristics of Holistic Data Analytics

Data does not speak for itself.

Every piece of data has a history of its own that explains how and why it is being collected, processed, analysed, and presented to the audience as a specific narrative. Holistic Data Analytics is an approach to data analytics that takes into account the end-to-end lifetime of the data.

Holistic Data Analytics has six characteristics.

Problem-oriented 

Holistic Data Analytics starts with a business problem to be solved, a question to be investigated, or new opportunities to be found: it is problem-oriented. This distinguishes it from the traditional approach that focuses on developing Business as Usual (BAU) reports. In the holistic approach, the analyst must collect, clean, model, and finally extract insights from data based on a central ‘problem’ of the organisation. This, in turn, requires the data analyst to have a deep understanding of the business, its processes, and its revenue models. This takes us to the second feature. 

Whole-of-business approach

A major pitfall in many practices of data analytics is that analysts often do not have a full grasp of the organisation, the meanings of the data they are working with, and the context in which data are collected and processed. This, often, delinks data analytics projects from the real business context: fancy charts and visualisations that rarely trigger data-driven decisions in the organisation and myriads of dashboards that do not talk to the realities on the ground.

The human context

Holistic Data Analytics pays particular attention to the overall lifecycle of the data, the context in which the data are produced or collected, processed, and analyzed, and how these analyses are being consumed by the stakeholders. More often than not, humans are the ones who collect, process, manipulate, interpret and consume data, and to solve a real-world problem with data, we often need to understand and analyze the human context of data collection and consumption. To verify the reliability of the data, analysts need to personally experience the operational context of data collection; sometimes they need to travel with technicians to sites to see how they fill out forms and collect data; they need to observe clients’ behavior to understand their journey in their online purchase, and they need to seat with admin staff to examine how they process and act on data. These observations of human behavior are central to how the analyst selects specific data fields at the expense of other data fields, how she models and integrates those data, how she attributes weights to each field in the analysis, and how she communicates the resulting insights with the stakeholders. 

To solve a real-world problem with data, we often need to understand and analyse the human context of data collection and consumption.

Analysts need social skills

As a prerequisite, a data analyst needs to have a mastery of diverse technical skills: statistics, mathematics, analytics software and services, and programming languages. Going beyond the surface of data to provide valuable and actionable insights, however, requires more than computing, coding, and numbers skills. Analysts must have people skills: effective verbal communication, teamwork, and social and cultural literacy. To achieve even a higher-level interpretation of the data, analysts need to have a decent grasp of human behavior, social psychology, organisational behavior, and communication. 

Every data story highlights some facts and hides others 

Every narrative, including data narratives, highlights some facts and hides some others. 

For traditional analysts, available data are given truths and gold-standard facts, and no question beyond the very surface of the data must be asked. In Holistic Data Analytics, in contrast, every piece of data has a story and a narrative that highlights some facts and hides some others. Telling a story with data inevitably means selecting some pieces of data and making them visible at the expense of some other pieces. Most data analysts, especially in big organisations, may have experienced the puzzling moment of having to select a handful of columns of data from among hundreds of columns. Often made behind the scenes, these decisions foreground some data fields making them visible to stakeholders and downplay the role of other data fields in the overall analysis. Every narrative, including data narratives, highlights some facts and hides some others. 

Data that do not exist are as important as existing data

Be vigilant about data blind spots 

Even with implementing the best analytical approaches, sometimes available data will not yield useful insight for the business problems; they will only lead to the same old results. Determining what data points to collect requires some assumptions, narratives, and theories about the working of the organisation, customer behavior, and the economy at large. Often, these naïve theories are not verbalised but are unconscious and we act upon them. They determine what data points to collect and what data points to ignore as irrelevant. Once collected, any insight based on these data will often validate those old assumptions. That’s why being vigilant about the data blind spots of organisations is as important for solving business problems as analysing the existing data.

AxiomNetic's holistic data analytics is designed to provide problem-oriented insights that encompass the entirety of your business and its human context.

Contact us for a free initial consultation.