Parcel Perform Uses AWS to Boost Customer Lifetime Value by 40%

Logistics can be a major headache for ecommerce merchants. To deliver packages to a widespread customer base, merchants in Singapore often work with as many as 10 different carriers, such as DHL or Singapore Post. Each has its own backend system, which ecommerce merchants must integrate with their own operations. This involves significant time and resources.

Dr. Arne Jeroschewski spent more than a decade in senior management roles in ecommerce logistics companies and experienced these challenges firsthand. In 2016, he founded Parcel Perform in Singapore to improve visibility and efficiency in the ecommerce supply chain by centralizing logistics management. The Parcel Perform platform is a global, enterprise-scale, parcel-tracking software as a service (SaaS) solution that promises its merchants a 40 percent increase in customer lifetime value and retention. The increase comes from a combination of factors, including lower servicing cost per customer and the ability for customers to independently track parcels without contacting service agents.

At the core of its product offering, Parcel Perform aggregates and standardizes logistics data from over 600 carriers worldwide into easy-to-understand tracking updates. Based on this data, the company then offers its customers fully branded, white-labeled tracking pages and automated customer notifications to improve visibility for ecommerce merchants and their consumers.

Parcel Perform was born in the Amazon Web Services (AWS) Cloud. The company’s founders had experience working with AWS and recognized the value of its continuous pipeline of innovative technology solutions. “We felt comfortable that we were in good hands for continually evolving our business, even without knowing what technologies we would require in the future,” Dr. Arne Jeroschewski says. “There are many moving parts in our hosting environment and AWS has proven to be flexible, scalable, and supportive of our business growth.”

Parcel Perform now handles hundreds of millions of shipments a year and works with over 600 logistics carriers. Data processing and integration are at the core of its operations. The company uses Amazon Relational Database Service (Amazon RDS) to automate database administration and Amazon ElastiCache for short-term storage caching with low-latency data retrieval.

As the business grew, it introduced container-based architecture. While parcel tracking is predictably a daytime function, infrastructure size requirements can vary widely due to differences in workloads across multiple time zones. During peak shopping seasons, use of the SaaS solution may spike by 80 percent. With Amazon Elastic Container Service (Amazon ECS), Parcel Perform’s hosting environment can scale according to demand.

For successful software companies like Parcel Perform, cost optimization and performance usually go hand in hand. The business uses a carefully calibrated mix of Amazon EC2 Reserved Instances, On-Demand Instances, and Amazon EC2 Spot Instances to give it the flexibility needed to quickly scale its infrastructure during peak periods such as Christmas and to shut down instances when demand falls.

“By using tools such as Amazon ECS and Amazon EC2 Spot Instances, we can hit just the right level of provisioning to deliver the service we promise to our customers without going over budget,” relates Dr. Arne Jeroschewski.

According to Dr. Arne Jeroschewski, about half of all customer service calls to ecommerce agents are to check on a parcel’s location. Merchants using Parcel Perform benefit from proactive, merchant-branded status updates sent to their customers via email or SMS communicating where packages are and their expected delivery date. This substantially reduces the burden on customer service agents, who can redirect their efforts to more value-added tasks.

For most ecommerce delivery journeys, there is a period of awkward silence in customer communications from the time the parcel is lodged until it is out for delivery. Parcel Perform recognized this as an important stage in the post-purchase experience and ventured into the realm of machine learning (ML) to leverage its vast amount of data to create a delivery prediction model.

The company started its ML journey by working closely with the Amazon Machine Learning Solutions Lab, using Amazon SageMaker to build and train its prediction algorithm. With this algorithm in place, Parcel Perform can bridge communications and provide tailored parcel updates with accurate predictions for date of arrival. “We appreciate the technical input from the AWS team, and it was helpful to combine that with our domain expertise on the logistics side to come up with this function,” Dr. Arne Jeroschewski says.

The Parcel Perform service goes beyond predictive capabilities, offering merchants retrospective analytics with its logistical intelligence suit, which is essentially a business intelligence suite. Engineers have built a data pipeline from their transactional database and Amazon RDS into an Amazon Redshift data warehouse. From there, they use visualization tools to give merchants a side-by-side, at-a-glance comparison of how well different carriers are performing.

Using dashboards and a host of graphs and tables included with Parcel Perform, merchants can view actual transit time and deliverability ratios per carrier. This visibility empowers merchants to make data-driven decisions, keep their logistics partners accountable to their SLAs, and optimize logistics spend.

Normally, Dr. Arne Jeroschewski explains, logistics companies don’t provide granular data on their KPIs, but with Parcel Perform, merchants can manage their partner ecosystem more professionally. Carriers are now held to higher standards. By choosing carrier partners that provide more value for the money, merchants benefit by improving the service experience for customers while also lowering costs.

Parcel Perform has stayed loyal to AWS because of the support received, not just from a technology perspective, but also from a business standpoint. The company is currently exploring new go-to-market opportunities. Dr. Arne Jeroschewski concludes, “What really stands out for us with AWS is that it is visibly committed to our business success. Our AWS representatives are always available, and over time we’ve come to appreciate the continuous support we receive as an AWS customer.”

To learn more, visit Machine Learning on AWS.