INVENTORY DATA ANALYTICS
Inventory data is a gold mine for information based on which organisations can make critical business decisions. The available data can often be quite huge and it has to be formatted and presented in a way decision makers can make sense of it. This is where data analysis comes into play. Data analysis essentially extracts useful insights into customer behaviour, product movement and other important indicators from the vast amount of data.
Elementary Analysis
An elementary analysis on the data itself can reveal a lot of useful information about sales trends and customer behaviour. The above plot shows the profit margins from top selling products of the company over the years. Rice, Sugar and Edible oil remain the top three grossers. Products like home care, packaged foods etc are starting to contribute more in 2020 mainly due to people being confined to their homes during the pandemic.
The two plots here,reveal a very interesting story. The first plot shows the sales trend and margins over the last couple of years. the sales and margin have been steadily increasing until March 2020 after which there is a significant dip in sales. The dip in sales coincides with the start of country wide lock down in India due to the Corvid-19 pandemic situation. But the story doesn’t end there.
The second chart reveals that the sales of rice has gone down during the lockdown period. Rice is one thing that people would like to stock during uncertain periods like these, but still its sales has dropped, why? The reason for this cannot be found with the data that we have in hand. A closer look at other sources brings to light the fact that government has issued free ration of rice to all the people of the state during lockdown and it has impacted the sales figures of rice. This also highlights the need to correlate other sources like new articles with the data to bring out facts we otherwise would miss.
Suggested Action Points
- To overcome the sales dip due to lockdown, the company should give more focus on products that showing growing customer interest like homeware, packaged food etc
- The company should gather information from news and other external sources to forecast trends and initiate strategies to counter changes.
Recency, Frequency, Monetary Value Analysis
RFM analysis provides many details about customer behaviour and it facilitates targeted action for different kinds of customers. RFM analysis is build around a customer’s recency of purchase (R), frequency of purchase (F)and gross monetary value of his purchases (M). Based on the values of R. F and M customers can be divided into different classes and a separate set of actions can be advised for each of these classes.
The below figure shows the customers based on RFM analysis on the sales data.
Champions
These are the best customers and they have purchased recently, purchases very frequently and spends a lot of money on their purchases. It is very important to hold on to champion customers through extension of gifts, exclusive offers or other means. In this particular case only 22% of the customers are champions and the company needs to find ways not only in retaining these customers but also to bring more customers as champions.
Hibernating
Hibernating customers are those customers who were champions or loyals but has not purchased recently. If not properly attended the company very soon loose those customers.Factors that lead to hibernate must be eliminated by taking extra note on this categories. 23% of the customers are currently hibernating and the company needs to pay urgent attention.
Loyals
Loyal customers purchase very frequently and if given proper attention could be converted into champions. Here 10% of the customers are loyal.
At Risk
These are customers the company has almost lost. Attemp can be made to bring them back.
Lost
These are the customers company has already lost. There is no point in investing too much into these customers but they can make an attempt to bring them back.
Potential Loyals
With proper attention they could become loyal and later champions. 10% of the customers are potential loyal.
Suggested Action Points
- Around 44% of the customers belong to either champions, loyal or potential loyal customers. It is important for the company to make sure that these 44% remain as customers.
- It is also important to bring back the 23% hibernating customers which could make a potential 67% champion/loyal customers.
Technology road-map
Data analytics can reveal a lot of hidden information about the business and can lead to action plans for more efficient practices in the future. Besides these insights and action plans we also advice a detailed technology road map which would let the company execute the action plans in the most productive manner. Technology road map include the execution of the following tools.
Logistic management tool
The company could make use of a logistic management tool which would save them cost and improve customer satisfaction. This tool would automate all the logistic tasks and at the same time will make them more efficient. The tool would optimise routing and clustering of transport vehicles, sort and allocate shipments, track fleets and ensures timely delivery of goods to customers
Direct selling through WhatsApp
The company follows a B2B model and this meant that they were even more heavily effected by the lockdown. But they could find opportunity in this crisis by a shift to B2C model through direct selling. They already have the infra and logistics and all they need is a platform. Direct selling through WhatsApp is the best possible scenario for the company as it is easy to implement and most convenient for the customers.
Recommendation system
The RFM analysis and market-basket analysis would provide a Highly effective Recommendation system and good insights into customer behaviour. These inputs, with the help of an automated recommendation system, could help to suggest new products and offers for the customers depending on their class of purchase history. This would help the company in customer retention and upgradation. With the aid of srong recommendation system implementation champion customers could be rewarded. The loyal/potential loyal customers then could become champions and lost/almost lost customers could be brought back into the fold.
NLP to correlate external sources
Sales trends can be largely influenced by many external factors like pandemics, political scenarios, clarities and weather. It is important for business to keep track of these external influences and proactively react to them. Using NLP techniques, news articles can be regularly analysed to extract useful information about upcoming events that could have an impact on the sales. This information would then help the decision makers to come up with solutions even before the event would have an impact on their business. For example previously we have seen how government's free ration scheme has impacted rice sales. With NLP, the company could have made preparations to minimise its impact on their sales.
Anticipating issues beforehand
ML can help ERP systems anticipate problems before they occur. Such advance preparation is only possible with ERP software that uses artificial Intelligence.ERP Systems with Al will be able to demand forecast,avoiding wastage of resources and preventing shutdowns as well.
