Want to improve your Kafka projects cost efficiency & performance? Download our free eBook

Major Retailer Gains Visibility into Orders During Peak Season

Helps to ensure the ‘price is right’ by drastically improving efficiency and cost savings during peak retail season.

Business Transaction Management for Major International Retailer

The retailer’s thousands of unused queues allowed it to consolidate several unneeded servers, providing annual savings in excess of $400,000.

The Problem

For a consumer retail company, the period between Thanksgiving and New Year’s provides both great opportunity and great risk. meshIQ worked with a retailer that earned more than half of its sales revenue during this one time of year, and experienced massive volume that stretched various system capacities. Companies that have both online and offline retail channels experience additional difficulties in managing inventory and ensuring consistent pricing, highlighting their need for process visibility.

Overview

Interconnected Problems Online and Off

Various processes and systems within the retailer’s brick and mortar and internet sales channels were not working together at optimal levels. Their process for pricing involved a data push from their headquarters to all of their retail stores. The pricing information also needed to go the point of sale (POS) systems in each store and to the specialized system that created pricing labels for the aisles. Furthermore, the pricing updates were sent to a third-party who printed circulars with sale prices. Discounts and short-term sales were especially troublesome to track and coordinate. Unfortunately, the pricing update transactions would sometimes fail for some of these locations, and the retailer had no visibility into these transaction failures. Their pricing processes dictated that all pricing transaction updates must complete or none of them would complete. The retailer did not have the proper business transaction monitoring in place to detect and enforce this policy.

A lack of visibility into pricing update problems led to incorrect pric­ing which drove order abandonment and customer attrition. Pricing inconsistencies produced loss of sales as consumers grew frustrated when presented with internet or paper flyer prices that did not match to in-store pricing. Many consumers would select a product and intend to checkout, but would then be alerted to a higher price and abandon the order at POS. Such in­stances placed a significant strain on customer loyalty and repeat buying behavior.

As this pricing inconsistency prompted cart abandonment by both the online and brick and mortar shoppers, inventory levels increased and were more difficult to manage. Forecasting was inaccurate. As carrying costs rose, the retailer began to lose revenue and customers and needed to perform additional work to adjust sales forecasts and other business metrics.

After an extensive search for the right partner, the retailer chose meshIQ’s AutoPilot solution for its proven track record of providing proactive deep visibility into business transaction, middleware, and application problems.

meshIQ began by providing business transaction management for the retailer’s pricing transactions in order to catch any failures in real-time. Next, it looked at the retailer’s middleware messaging where I discovered tens of thousands of queue objects in the retailer’s middleware management estate. The next step in the process was a comprehensive review of the performance data for each of these objects over a 25-day time span.

The results were enlightening. A full 26-percent of the queue objects that were deployed were no longer used by the retailer. Eliminating these objects created a more simplified architecture, which allowed meshIQ and the retailer more insight into the remaining queues. They used sub­stantial hardware to manage these unnecessary queues, producing overcapacity in some areas and under capacity in others. Our data analysis found 60-percent of the objects were configured incorrectly, including those that checked pricing, confirmed new products, and performed other logistical tasks. What the retailer and its IT department needed most was visibility. They were capable of fixing problems once they were exposed, and knew something was wrong in their processes, but had no view into which links in the chain were broken.

Improved Efficiency and Cost Savings

By utilizing meshIQ’s AutoPilot business transaction management and deep middleware monitoring, the retailer was able to quickly correct and streamline exposed issues. As a result of these fixes, order fallout was reduced by an astonishing 40% due to the elimination of pricing discrepancies, better stock fulfillment, and a smoother checkout process. Point of sale abandonment was decreased by 35%, reflecting the shifting attitudes of consumers who were no longer frustrated by pricing inconsistencies.

The retailer’s thousands of unused queues allowed it to consolidate several unneeded servers, providing annual savings in excess of $400,000.

Longer-Term Payoffs

After usage of AutoPilot’s real-time monitoring and built-in complex event processing, the retailer gained valuable visibility into order and fulfillment processes, allowing what were previously “stealth” problems to be preempted and new efficiencies realized.

Sales during the busy holiday season were no longer a source of grave concern for the retailer’s management and IT staff. Their systems were more reliable during the sales rush, saving millions through decreased cart abandonment and an overall improved customer experience.

The Result

Prior to AutoPilot deployment: They had a very complex, manually created map of business processes and transactions… difficult to maintain and monitor.

After AutoPilot deployment: The map of business processes and transactions was automatically discovered, monitored and measured.