Impact of Pandemic on Supply Chain  

The Covid-19 pandemic has affected normal living and has affected most of the businesses including retail. During the unprecedented time, businesses took extraordinary efforts to maintain the supply chain to its full effectiveness. In this blog, we will focus on how one of the UK based retail clients dealt with the situation to minimise the impact on the fulfilment of daily customer orders by reimagining the existing supply chain solution/capabilities and how the Performance element was thought through while doing so.  

Client Introduction and Customer shopping behaviour (before Covid-19)  

The client is one of the largest Pharmacy led Health & Beauty Retail business in the United Kingdom.  With the widespread presence of their business across the UK, the majority of the revenue comes in through the brick & mortar channel while the online channel’s contribution has been in the low teens. For the online orders, the preferred delivery option was ‘Collect from Stores’- which accounted for more than 65% of the online orders.  

Impact of Pandemic on Customer’s shopping behaviour  

As a response to Coronavirus Pandemic, In March 2020, Government announced countrywide lockdown across the UK. This resulted in drastic changes in customer shopping behaviour contributing to an unprecedented surge in customers preferring online shopping over in-store shopping. Even within the online shopping, customers specifically preferred ‘Home Delivery’ orders as opposed to ‘Collect from Stores’ orders. The Client’s warehouse was overwhelmed with ‘Home Delivery’ orders and owing to the pandemic, staff attendance was also diminishing due to isolation. In short, the warehouse was struggling to meet the delivery promise date.  To help the warehouse, during the initial period of lockdown, several operational changes were made like extending delivery lead time, limiting the number of customers on order placing source application i.e. Client’s e-commerce site, etc. This resulted in the loss of a large number of orders.  

Solution – Re-Imagine/ Re-Purpose/ Re-Think  

To overcome this situation, several ideas/thought processes were brought on to the table. The aim was to fasten the order processing without impacting customer. Rapid Action Team (RAT) was built to help repurpose existing Order Management Solution in such a way that stores can process ‘Home Delivery’ orders to ease the pressure on the warehouse. The ‘Collect from store’ fulfilment process was repurposed for ‘Home delivery’ orders with minimal changes. Several Big stores were identified which then were converted to Hybrid stores/Tiny Warehouse i.e. stores that can serve ‘Home Delivery’ Orders as well.  Performance Assurance was integral part of the RAT and played key role in decision making as this change has impacted number of systems from performance perspective. 

‘System Performance’ – Key element for Re-Imagined Solution  

Major challenge was the impact on Performance as all Home Delivery orders were converted in to Collect from Store orders. Traffic for Collect from Store orders were 4 times greater than for Home Delivery Order. 

Appropriate test suit/framework was built using PT tools like NeoLoad and JMeter using which each design change and impacted systems were performance tested quickly. This has helped in experimenting changes and analysing impact in short time. Tools like Dynatrace, Insights, Mongo Portal, LogicMonitor, NMon were used for Monitoring and analysis. Key performance improvements made are as below:

• Processing time taken by an order was improved significantly. During the initial PT run, time observed was 18 minutes which then was improved to 12 minutes.
• Optimum number and CPU/MEM config of Pods for each micro service was identified.
• Optimum config for Mongo Atlas cluster config was recommended with higher IOPS/throughput rate and with higher number of connections allowed.
• Mongo Atlas DB CPU utilization during evening and mid night heavy batch processing was optimised. During initial phase it was observed to be 85% which was then optimised to 40%. 
• For drastic increase in number of records in DB, Archiving solution was recommended.

Roll Out Approach 

Phased rollout of this Reimagined solution was recommended. The approach taken was to try out the solution for one store, get instant feedback, and then move forward. Also Switch off/back out option was recommended, and performance tested in case of failure. From the inception, solution was implemented to first store in a week, then by increasing number of stores per week, solution was implemented in 150 stores in 3 months. 

Benefits

By implementing the solution for Hybrid Stores, the client was able to cater to the 180% increase in online orders. Almost 40% of the online orders were being fulfilled by the Hybrid stores.  Since the launch of the solution in early April, till mid of Dec 2020, all hybrid stores/tiny warehouses have picked most number of orders and contributed to revenue in large manner.

During second lockdown announced in early November 2020, Over the course of 4 days of peak period trading period, significant number of orders were picked by solution which caters to 60% of total orders placed. The solution helped in achieving biggest ever trading period for the client in recent years and even with this heavy load, solution was stable.

Conclusion  

In this journey Performance Assurance team showed proactiveness and solution is already proven for future state i.e. can be implemented in twice the number of stores without an issue as opposed to the number of stores for which it has been implemented now.  Another big advantage of this solution is, although the solution has been created out of necessity, the solution could potentially be utilised as long-term order fulfilment strategy for the client.  So, to put it in simple words, by rapidly repurposing the existing supply chain solution/ capability, the client has maintained their order fulfilment rate to its full effectiveness and have dealt with the risk of losing customer orders effectively during the lockdown.