Apr 25 2017 | Analytics
Use of analytics in traditional industry
By: Isha Mahapatro, Enterprise Risk & Int Audit

We all have heard about ‘Data Analytics’ which is changing the way an organization works.
Today Data Analytics has found immense application in many fields and disciplines like:

  1. Finance
  2. Marketing & Sales
  3. Operations
  4. HR

With the humongous amount of data traces generated by individuals and organizations and the ability to store the data at a lower cost has led to a revolution in the science of analysis and decision making.

Overview of type of Analytics:

  1. Descriptive: It says what is happening based on the data currently available with us. A real-time Dashboard/Graphs/Charts could be best to understand the data at hand and identify patterns or insight to make decisions.
  2. Predictive: What might happen in future based on the current data available. Also it includes predicting the outcome of an event (dependant variable) based on independent variable influencing the outcome.
  3. Prescriptive: It talks about the action to be taken once we predict a scenario. Like doctor prescribing specific cancer medicines to patient who he thinks might develop cancer in future.

Here we will try to understand on how data analytics can benefit our traditional sector and improve its performance through a case of textile business.

Step by step implementation of Analytics:

  1. First step is to understand the business and underlying processes. This will help in understanding various type of data that can be gathered by a consultant at various stages of production and sale. Also we get a view of what can be automated for better performance and can act as a source of data. Example-
    • Introducing bar code for scanning finished clothes before shipment generates data.
    • CRM (ERP) system to capture various details about customer buying the products etc.
  2. Once the IT systems are in place and they can capture the day to day data on customers, sales, supply and distribution. Before using this data for analysis the most critical step is data clean-up. This includes:
    • Treating missing values or human errors, identifying outliers etc.
  3. Descriptive analysis of data to get a first-hand information:
    • Plotting a simple graph or chart can give us quick insight on something that is happening but went unnoticed. Example- Graph of sales against the months can give us a seasonal purchase pattern for a product.

    Other tools that can be used are Pie chart, Histogram, Box plots etc. Heat maps are another interesting way to analyse given data, excel add-ons can be used to create such maps:

  4. As the company has many products, it becomes essential to categorize these products based on their sales or some purchase pattern. Once done the various groups of product may follow different strategy to boost sales in an effective manner. Below is Rate (x-axis) Vs Sales(Y-axis) in Kanpur. Clearly 2 clusters, red and green make the best sales.

    Using tools like time slicers and visual filters beautiful dashboards can be created.

  5. Use of predictive analytics to forecast sales of cloth products: Initially a simple forecast can be done to predict month on month sales. Exponential smoothing or seasonal forecasting can be done. But soon you will realize the cloth sales (sarees and ethnic wear are event based) and gets impacted by Hindu events like wedding seasons , Diwali, Dussehra etc. which changes as per Hindu calendar hence an event based forecasting is the key to accurate forecasting. Statistical software like Forecast Pro could help in such forecasting.
  6. The Final step is to automate all this analysis which can be used by the management for critical decisions related to sales and production planning So, through this blog we saw a practical case of how analytics coupled with IT could revolutionize a traditional setup like textile organization.