Case 1
Purpose
The client wanted an all-inclusive business intelligence tool to help merchandisers make informed decisions
Approach
We connected with the merchandisers across different merchandising groups to understand the insights they draw from various weekly business reports
Using the insights the merchandisers need to gauge their weekly performance we backtracked the KPIs they need to track. The KPIs were put into a storyboard to understand the information flow across merchandisers and the type of decisions it would enable
The decision flow storyboard was then transformed into flat design views by our expert design team and a cloud based web dashboard was build that was rolled out across phases to different merchandisers.
Outcome
The clear decision flow of the tool enabled merchandisers to have a 360-view of their business area.
Merchandisers were able to save 20% of their valuable time which was earlier spent in looking at numbers across multiple reports and comparing them
They were easily able to draw clear business insights about their merchandising group and make rapid informed business decisions
Case 2
Purpose
The client, a small US based supermarket chain wanted to find the optimal product ranging strategy to improve product availability and customer satisfaction by 10% on a sustainable basis
Approach
We collected data from 5 data sources including data from external vendors and applied proprietary time series and machine learning algorithms to predict the impact of removing and adding products to a store
Initially we looked at 5 years of data and built time series models across all food categories
We then built individual product level contribution models using sophisticated machine learning techniques.
The contribution models were used along with substitution models to gauge the impact of adding or removing products across categories and stores
Outcome
Using the model the client was able to cut the existing range to 70% of the original. This resulted into cost to serve reduction by 15% and product availability improved by 20%
The un-bloated range resulted in improvement in customer satisfaction score by 8%
15% decrease in marketing expenses in introducing new products that didn't have potential to grow and improve customer experience
Case 3
Purpose
The client, a UK based retailer wanted to know the optimal promotion strategy to attract customers and improve sales
Approach
We aggregated data from multiple data sources to create a unified data mart. the unified data mart contained data of
- All product sales and volume
- The amount and type of promotions run on every product
- The seasonality effects across products
A unified market mixed model was built to segregate effect of overlapping promotions and calculate the price elasticity of different products. This price elasticity value was used to gauge the impact of promotion on sales of the products
Outcome
Our client was able to reduce its promotional expenses by more than 25% by targetting only the promotionally responsive segments of products
There was a improvement of 15% in the overall profits due to reduced costs and marketing expenses and the customer satisfaction grew by 5%