Subject: Special Offers Evaluation for a Retailer in Automotive, Sales Analytics Toll
Data Science Areas: Machine Learning, Data Processing, Predictive Analytics, Business Intelligence, Business Forecasting, Sales Forecasting
Tools: Python, Pandas, Numpy, Tkinter
Summary: For a dealer in the car parts retail business, a team of machine learning developers at MindCraft create a Predictive sales analytics tool. It automatically measures the profitability of buying items from the special offers lists. So, the solution evaluates the possible sales rates and estimates the discount. It provides invaluable marketing insights and helps to make smart business decisions.
The Challenge of sales analytics tool :
Dealing in the car parts business, our customer receives a lot of special offers, because in this industry, there are millions of items and a special offer document can contain tens of thousands of items. It is usually an xls file looking like this one:
If Such an offer may contain from a few hundred to tens of thousands of items with discounted price. Most of those positions rarely or never sell. For instance, there is a poor chance to sell a valve for BMW Z1 since only 8000 cars were ever sold. Also, each car has thousands of car parts, most of them work well for years.
The evaluation of such a proposal takes up to 1.5 hours and this work requires extremely high qualification. The number of special offers can vary from 1 to 7 a day. The correct evaluation of one such offer may result in thousands of dollars of revenue. If wrong, the company may spend thousands of dollars on items they will never sell.
This process was performed manually. An employee would check the sales report and compare it to the stock availability of top competitors.
So, the client asks MindCraft to create a predictive sales analytics tool based on machine learning. That would automatically check the potential sales rate of each item on the list. The results would be picked also with regard to the size of their price discount
While dealing with big data cases, like this one, our team of data scientists always strive for the cleanest data. Our data analysts analyzed all sales reports provided and proceeded with filtering the data. The following factors were taken into account:
- To achieve the highest data purity, we removed the wholesale car parts dealers who keep their own stock from the list
- We analyzed the average monthly number of items sold
- The tool estimated how lower the discounted price is comparing to the normal one and the minimal price of the item per year
- We calculated the number of clients who bought the items
With the help of a machine learning model, all these operations are now being performed automatically. There is no need to compare all these lists manually to the competitors’ data. What our clients now use on a daily basis is a user-friendly interface, which does everything for them and looks like this:
What comes out is a simple and clear document like this:
This machine learning tool doesn’t replace the professional employee with the knowledge of the industry. However, it takes about one-fifth of the time this employee would spend on all data lookups and comparisons. What was done manually using MS Excel and the accounting software – is now all in a simple and elegant tool. In fact, the predictive sales analytics method we used can be of a great benefit to any B2B or B2C retail business. With a little customization, this data-driven technology can help organizations quickly make the right marketing decisions.
Pay attention to the second part of the project: