Case Study

How a Digital Logistics Company Used Data to Rethink Operations


We worked with QuadX, the fastest growing digital logistics service in the Philippines, to develop a data management & visualization solution that unifies information from multiple parts of the company, streamlines reporting, and encourages employees to adopt a more data-driven mindset. Thinking Machines helped QuadX to:

  • Make decisions instantly based on real-time, validated data as opposed to 2 week-old, incomplete data.
  • Reduce time needed to prepare reports, from 3 days down to 2 hours.
  • Full access to locational and transactional insights for over 60 employees.

Schedule a consultation

The Challenge

Creating a Future-Proof Data Architecture

QuadX is a fast-growing logistics company, shipping tens of thousands of packages every day. To make the user experience as easy as possible for millions of customers, QuadX uses data and analytics to help drive operational decisions. This includes using data to obtain complete visibility on fulfillment times, identify underperforming stores, and pinpoint bottlenecks such as warehouses, point-of-pickup, or point-of-delivery.

In order to keep up with the demand, QuadX needed to create a data management system to reveal better insights and a visualization tool to put data in the hands of everyone in the business.

"Our competitive edge is dependent on delivering timely, low-cost services to our customers for which we need up-to-date, accurate data," says F.M. Ycasiano, business intelligence/data analyst at QuadX. "But we had no data warehouse, no operational dashboards, and no system that could handle all the information we were collecting. The biggest problem was that our operations reports were incomplete and already 2 weeks old by the time they got in the hands of management. We knew we could do better."

Another challenge: It was impossible to create a data-driven culture with faulty information. Without the proper data infrastructure, staff could not root out inherent bias from their decision-making process.

QuadX turned to Thinking Machines to develop an integrated solution that would enable fact-based decision-making and make data accessible to everyone within the business.

The Solution

Data Management & Visualization



Diagnostics

We dug into QuadX’s data related to its two main business lines, ShippingCart and CheckMeOut. This entailed interviewing internal users, evaluating existing systems, foundational data analysis, and answering key questions such as:
- What are the key critical operational challenges?
- What kind of data is available to support decision making?
- Where and how is this data stored?

Data Warehousing Design

QuadX used to perform manual queries with PostgreSQL to generate delivery and fulfillment reports. While that workflow worked well for simple, one-off queries, QuadX needed a service that was faster and more cost-effective. In addition to simple querying, the analysts wanted to combine data from different sources and show their reports on a dashboard, both of which underscored the need for a single data repository. After considering the pros and cons of Google BigQuery and Amazon Redshift, we decided on BigQuery because it was a fast and fully managed solution. This enabled QuadX to spend more time on analysis and less time on optimizing queries and maintenance.

A data pipeline was also needed to move data from their databases to their data warehouse. Some of the options we tried included AWS Data Pipeline and AWS Glue, Apache Airflow, and cron to schedule our ETL (extract, transform, load) jobs. We ultimately decided on a simple cron scheduler of ETL scripts written in Python, because QuadX placed heavy emphasis on maintainability and transparency of the pipeline.

 
Google BigQuery
Amazon Redshift
Fully managed?
Yes
No
SQL support?
Yes
Yes
Need to optimize indices?
No
Yes
Billing
per query +
storage pricing
Based on
no. of clusters



Data Visualization & Capacity Building

To make the analytics insights available to all non-data savvy stakeholders across the 60-person company, we connected the data warehouse to QuadX’s PowerBI platform. Our Data Visualization Workshop equipped their staff with the necessary skills to build their own customizable reports for business users at every level. Furthermore, we delivered consulting services to put in place sustainable processes around effective data management.

The Tools

Future-proofing the System

Thinking Machines worked with logistics analysts at QuadX to determine what tools would be most appropriate for their data ecosystem, especially given the need to incorporate additional pipelines down the road.

  • Google BigQuery: a cloud-ready data architecture on which multiple data sources could be stored

  • PostgreSQL: a database management system which enables users to request specific types of information

  • Microsoft PowerBI: a business analytics service that provides interactive visualizations and self-service business intelligence capabilities

"How we operate is changing the way QuadX works because we can see the numbers. With the data readily available we can be much more efficient."

Dino Araneta
CEO

The Impact

Laser-sharp Monitoring and Improved Fulfilment Times

After only two months of operation with its cloud-based data system, QuadX was able to see exceptional results. Analysts and business users now have order-level visibility across the company’s entire delivery chain, which was impossible using the previously siloed data system. As a result, QuadX’s leadership can better identify the root cause of complex problems and quickly respond to meet higher standards of excellence.

The data management & visualization solution has also helped deliver significant time savings, freeing up staff to perform more in-depth analysis and consider other functional areas that may benefit from a data-driven approach. The cloud-based system boosts employee productivity by streamlining the reporting efforts of 4 analysts down to 1 analyst and cutting time from days to a matter of 1-2 hours.

"When we got the data warehouse, all of the sudden, we could see across all our critical operational factors: supply chain partners, time periods, and geographies," says FM. "Now that our staff has seen the advantages, speed, and capabilities, they are not only immersing themselves in their own data, but also asking us to integrate their own data pipelines for more insight."

  • Make decisions instantly based on real-time, validated data as opposed to 2 week-old, incomplete data.
  • Reduce time needed to prepare reports, from 3 days down to 2 hours.
  • Full access to locational and transactional insights for over 60 employees.

Can we help you solve a similar problem?

Get a free consultation