OUR SOLUTION

Stage 1: Business Analysis

As with all our clients, we began by conducting a business analysis on its operations and strategy.

​

What were our initial observations and hypotheses?

​

• The Client has a small hardworking team that is often over-stretched.

• The Client had under-invested in process automation tools, resulting in an over-reliance on manual interventions in business processes that are slow and prone to error. This is very common in many SMEs in the traditional non-technology industries.

• The Client had under-invested in data analytics, resulting in an inability to rapid understand the latest business trends.

Stage 2: Data Collection

​

The company has an ERP system which has a lot of the data that is needed for analysis. For the data that is not in the database, we started collecting it.

​

Data Collection Steps:

1. Set up a SQL database to store the data not currently stored in the ERP system.

2. Decide on the data to be collected - type of data required, collection frequency, collection targets etc.

3. Where required, develop web-scraping programs to scrape publicly available data on various websites.

Data Required:

• Pricing data for the products that the Client distributes and for the competitor products.

• Sales and inventory data for sales and inventory analysis.

Stage 3: Data Analysis/Hypothesis Testing

​

We run the analysis on a regular basis to test our hypotheses.

​

Data Cleaning:

1. Use data visualization tools for preliminary data analysis on the database. This makes it easy to spot problems with the data, such as outliers due to errors in the data entered and collected.

2. Deal with outliers and missing data.

​

Data Analysis Methodology:

1. Run algorithms to check the integrity of the database. Where there are outliers or obvious discrepancies, we will work with management to resolve. This serves as a tool for Internal Control.

2. Use regression analysis to establish statistical relationships between key variables in our sales numbers to develop a sales model. We like using regression analysis at this stage because it makes relationships between the various components of our database easy to understand.

3. Once we have enough data collected, we will develop machine learning models for prediction purposes.

Stage 4: Change Proposals

​

Based on the analysis results, we develop proposals to revamp the business. The main objective is to modernize and automate many of the processes and integrate data analysis into the regular business decision-making.

​

Proposals:

1. Develop a model to analyze the effectiveness of promotions by analyzing the price elasticity of demand and cross elasticity of demand.

2. Develop a machine learning sales forecasting model to forecast sales and integrating the sales forecast with an inventory order management system.

3. Automate some of the manual processes, where applicable.

4. Develop a program for internal controls purposes, utilizing information from the internal databases.

5. Develop a set of tools for collecting and analyzing market intelligence, both soft and hard data. The market intelligence can be a useful competitive advantage for the Client.

Stage 5: Change Implementation

​

Change implementation For this particular Client, we work closely in the Change Implementation stage to implement the proposed changes to the business. This includes working closely on a daily basis with the Client to develop detailed implementation plans and schedules.