Assisting retail planners and buyers through a basket analysis decision appKey Benefits
Pepkor is a South African-based investment and holding company. It manages a portfolio of retail chains focused on the discount and value market. Over the years, the group successfully affirmed itself as a diversified retailer of significant size and scale trading in clothing, general merchandise, household goods, and cellular products and services.
One of the keystones in managing business activities across the Pepkor group is comprehensive data analytics. To achieve this, the group engages Pepkor Data & Analytics – a division dedicated to providing data science, analytics, and data engineering services. Likewise, Pepkor Data & Analytics works closely with the group’s business division by seeking answers to questions essential for making business decisions. In recent associations, the business division directed their concerns around basket analysis to Pepkor Data & Analytics.
The following case study considers the collaboration between Pepkor Data & Analytics and Spatialedge in developing the Basket Analysis Decision App (BADA).
Basket analysis overview
Basket Analysis involves analysing the purchasing patterns of consumers by observing the different combinations of goods shoppers buy together. This action allows retailers to quantify consumer behaviour - obtaining valuable insights for developing sales strategies.
There are quite a few examples of Basket Analysis in brick-and-mortar retail stores. The most important form of basket analysis is to compare the basket compositions (“baskets”) before and after changing product placements, store layout, or in-store displays. The captured data offers a way for analysers to deduce if customers purchased a particular combination at a single store or if the phenomenon occurs across all stores in the retail network.
The differentiating factor between standard sales information and a basket analysis tool is the practical insights retailers can gain. A basket analysis provides a deeper understanding of how consumers shop. When retailers apply this information in practice, it can result in an enhanced and more refined shopping experience for their customers.
In essence, a basket analysis tool enables retailers to drive initiatives such as changing the store layout or bundling products, understanding customers' behaviour and responding to it and accurately focusing on initiatives that improve customers' behaviour.
The Business Challenge
The Pepkor group consists of three main divisions: PEP stores, Ackermans, and the Pepkor Speciality division. These divisions have very similar analytical needs, including analysing baskets. Previously, PEP and Ackermans conducted their basket analyses using a combination of practices and tools, allowing them to investigate issues based on stores’ point of sales- or transactional data. However, the various tools and practices were costly and difficult to apply, making it a non-viable option for long-term use.
Moreover, experienced planners found onboarding new users challenging. Learner users were heavily reliant on power users for assistance even after several training sessions. This severely slowed down the speed at which analyses could be conducted.
Ultimately, the group called for a tool that aligns with their specific needs. The new tool had to be easy to use and offer flexibility for incorporating predictive and prescriptive analytics. Other requirements included merging the brand's current operating procedures in the functions of the new tool. Lastly, the group called for a single tool with the ability to answer a diverse range of questions regarding customer purchasing behaviours.
Considering the identified challenges and needs of the brands, Pepkor Data & Analytics initiated the development process for a new basket analysis tool. They required a team to assist with the implementation. Having worked with Pepkor Data & Analytics on numerous projects, Spatialedge secured our involvement based on our past deliverance of quality service. In addition, by leveraging off data engineering work and other tooling initiatives that our company has already provided to Pepkor Data & Analytics in past projects, we could fast-track the process.
Pepkor D&A and Spatialedge’s Approach
The project required us to include three additional developers in the existing team. We aligned our scrum architecture with the new development to ensure a constant frequency of sprint reviews and customer feedback sessions. As a result, we were able to develop a minimum viable product that satisfied the requirements of PEP stores, Ackermans, and the Pepkor Speciality division while affording them the ability to continue work and to stay productive as the development progressed.
We responded positively to the request from Pepkor D&A and increased the existing team with three additional developers. We aligned our scrum architecture with the new initiative to ensure a constant frequency of sprint reviews and customer feedback sessions.
“What impressed me most was the instinct the team had for what the user [us] wanted. Whatever requirements were sent they understood it well. And if they did not get it right or understood, they proactively sourced input from the business to get it right. Allowing us to get the interpretation right as well” - Rene Watson, Commercial Analyst, PEP
“...the knowledge on your team is amazing and the ability to understand the business requirements is amazing. We try our best to relay what the commercial side wants. And I think I've had experience in the past where we often don't get that right and then we get something that works from a developer side but sucks for the end-user. The responsiveness and the ability for you to make small frequent changes. That for me was massive.” - Duren Pachappa, Retail Analyst, Ackermans
After successfully implementing the new basket analysing decision app, our team compared the results with those of the previous tool. A summary of the results is depicted in the table below.
|Before: Various tools and ad-hoc processes||After: Spatialedge Basket Analysis Decision App|
|Overly complex systems making onboarding difficult. Only a select number of planners can use the tool.||A streamlined, intuitive user interface allows a seamless onboarding process with minimal disruptions, increasing the number of planners able to use the app the first time around without exhaustive training sessions.|
|A static view of the data is typically only available every two weeks.||Fresh raw data is available daily. Additionally, missing data is automatically populated and added by employing robust data pipelines, creating more trust in the data and the powerful ability to respond faster.|
|Execution time for running a single analysis could take anywhere between 15 and 60 minutes.||Execution time for most analyses is under 30 seconds.|
Pepkor Data & Analytics and Spatialedge are proud that the new Basket Analysis Decision App currently forms part of Pep stores, Ackermans, Shoecity, Tekkie Town, Dunns, and Refinery’s daily operations. Having data available within 24 hours allows planners to react to events that require immediate attention. Other daily uses include planning, post-mortems, business cases, proposals, and providing a realistic view to the executive committee.
A notable advantage of the new tool is the improvement in planning efficiency. Information is gathered quickly, resulting in proposals getting drafted in much shorter time frames. Planners can now promptly qualify proposals coming from store managers, thereby removing a lot of extra work. In this fashion, planners directly refer to the results delivered by the app to make data-driven recommendations rather than relying on the assistance of analysts or data scientists.
In closing, the key benefits of implementing the Spatialedge Basket Analysis Decision App centre on improving the user experience for planners, delivering fresh data, accelerating the onboarding process, and running faster analyses. If you believe your business can benefit from these features, contact us for a consultation.Download PDF