Making sense out of data

A crucial element in the data analysis process is effective utilization. This means making sense out of data. On average, 60% to 73% of all data within an enterprise is not used for analytical purposes. Most businesses have access to tons of data but are incapable of accurately analyzing and synthesizing it even though data analysis gives businesses an edge over competitors by improving decision making, service delivery and internal processes.

Data analysis (sometimes referred to as big data analytics) is the process of reviewing large amounts of data to unearth patterns or insights that allow for advantageous business development. For example, a dry-cleaning service provider might want to gain insights into which of their services appeal the most to customers. A survey can be conducted, and insights gained will drive the decision-making process in providing more appealing services to customers.
Data can be analyzed using three methods:

Descriptive Analytics
It is estimated that about 80% of business analytics is descriptive as this involves summarising and converting data into easily understood forms. Once this is done, patterns can be identified with useful insights synthesized. However, it is limited to representing data in tabular and graphical forms. Once the data is analysed, the process of drawing insights is left for the analyst to handle.
This form of data analytics utilizes historical information which helps companies in gaining insights into what has happened in a bid to make better decisions in the future. It is the easiest and most utilized form of analytics for businesses. For example, Netflix uses descriptive analytics to find correlations among different movies that subscribers view, this improves their recommendations engine. To achieve this, they use viewing history, historic sales and customer data.

Predictive Analytics
This is a more complex form of analytics that weighs probabilities to make forecasts about the likelihood of future outcomes. It answers the question of ‘what might happen?’ This form of analytics relies heavily on technology and an advanced skill set. A survey by Gardner Research revealed that only 3% of companies in the world do predictive analytics. A common application of predictive analytics is its use to produce credit scores in some developed countries. These scores are used by financial institutions to determine the probability of customers making future credit payments on time.
Also, marketing decisions can be backed by predictive analytics. In email marketing, content is developed based on insights gained from analyzing customers’ historic purchases, demographics and behavioural data. From these insights, decisions as to what content to include, how to re-engage customers, the most effective price points, what product recommendations and promo offers to dispatch are made.

Prescriptive Analytics
This builds on descriptive and predictive analytics to take data analysis to a deeper level. While predictive analytics tells businesses about the probability of future outcomes, prescriptive analytics goes a step further to make recommendations after weighing the pros and cons of different substitutes. Prescriptive analytics showcases viable solutions to problems and the impact of current choices on future trends. An example is Waymo, Alphabet Inc’s self-driving car. The vehicle analyzes the environment and decides the direction to take based on data. Just like a human driver, it decides whether to slow down or speed up, to switch lanes or not, to stay on course or take a shortcut to avoid traffic, etc.

As businesses in developed economies begin to embed more advanced forms of data analytics in their decision-making, African entrepreneurs and businesses need to ride this wave early because data analytics will help companies remain competitive in the global economy. African businesses must also progress from basic data analytics like descriptive analytics to more advanced methods.

We can start by implementing a data-driven culture. This will require the formulation of a data strategy that clearly articulates where a business wants to be in terms of its data analysis and what it needs to do to get there. This strategy will be broken into set goals for each sub-unit of the organization.
Specifically, the strategy should include ways that data gathered from different collection points in the organization can be centralized and made available to employees for analysis, including what technologies to be applied for different categories of data, and how employees can be equipped with the right skill set to handle the business intelligence function.