Clustering Application for

Pricing Analysts

About the Project

This project aimed at building a custom solution to aid the process of Clustering for pricing analysts to execute pricing strategies.

What is Clustering?

In a dynamic pricing environment, pricing analysts need capability to do ad-hoc deeper analysis of the competitive landscape to execute pricing strategies.


Clustering is an algorithmic approach to creating groups informed by key data points like store locations, proximity, drive time, competitor prices etc. The groups are then used to gain insight, enable quick analysis and deliver on pricing strategies in a nimble way in an omni - channel environment.

Timeline: 3 Months

Client: US-based Retailer (Name can't be disclosed for NDA purposes)

My Role: UX Designer

Team: Project Manager, Product Owner, Business Analyst, UX Designer, Five Developers

Design Challenge

One of the major challenge of the project was to advance the Clustering application towards user adoption as the initial phase of development focused on the requirements gathered to get the algorithms right and then eventually go towards user adoption.

Factors that hampered the overall user experience:

Performance Issues

User flows debts

UI Debts

Data/Calculation Inconsistency

Continous Discovery and Design

My role as a Designer was to redesign application to make it easy to use for pricing analysts to actively start using the application.

Firstly, the user stories from the current UI debts were added to the product backlog to be prioritized and built. We also took user feedback into account to further evolve this list of UI debts.


Secondly, some features were redesigned based on user inputs and priority. The designed features were tested with users to be refined and added as user stories to the product backlog to be prioritized and built.


Notes from the initial User interview/feedback sessions


Store Visualisation

Store visualisation helps pricing  analysts to visualise retail stores and competitor stores while applying filters for analysis before store linking.


To select the Retailer stores to be visualised as 'Reference Store'

To select the Competitor stores to be visualised as 'Reference Store'

To select the Competitor stores to be visualised as a 'Selection Stores'

Store Linking

Store Linking helps pricing  analysts to link each retail store with some competitor stores based on criterias like drive time, distance and location etc. These store links help in further analysis for the  Store clustering process.


Impact and Reflection

The pricing analysts actively started using the application for visualising stores for initial analysis. For linking process, we delivered some features but there was still some dependency on 'Linking settings' which were another set of features,  we planned and prioritised but did not deliver.


This project provided with an opportunity for understanding the retail domain. Also, working as an off-shore team was the most challenging part, in terms of communication with the users and coordination with our on-shore counterparts. The team did face some issues with respect to delivery but we were able to close loops while finishing the project.