© AdobeStock

Point of View

Transforming Customs operations management teams into strategic players

25 February 2021
By Customs4trade

Within companies, Customs operations management teams are typically slow to adopt new technology. Yet by adopting trade software solutions, they could centralize and automate their Customs and trade compliance processes and harness the value of data, saving time and money and informing strategic decisions. Although this article mainly addresses companies, it is also of interest to Customs administrations who want to understand the solutions and tools that are available on the market and that enable companies to modernize their Customs and trade compliance processes.

Data has become an integral part of all aspects of our lives. It is collected wherever we go, from the websites we visit to the loyalty programs we join, the apps we use, and the messages we send. It is then analyzed and used by business owners to optimize the effectiveness of marketing, sales, and operations. However, in order for the data to be useful, it must be structured – meaning it must be organized or stored in a pre-defined format. Structured data can be easily searched, allowing companies to perform analytics or obtain insights that would identify room for improvement, predict trends, and detect errors – the things that today’s companies truly need to remain competitive.

Customs data – are you in control?

As a company, if you are in control of the data required for Customs compliance, you are ahead of the game. But often compliance data is unstructured, sprinkled across internal systems as well as across ancillary systems belonging to Customs brokers, Customs authorities, community service providers, and port community systems. These systems are often not integrated, making data collection cumbersome and time consuming. Historically, this has held Customs operations managers back, prohibiting them from obtaining a full overview of their operations.

Structuring data through collection and centralization is game-changing. Data can be analyzed to reveal insights into metrics and KPIs that can enable growth and inform decisions on sales, procurement, the supply chain, operations, and more. It can be fed back into other source systems, for example ERP[1] systems or WMSs[2], making each system smarter.

While other departments within a company have been reaping the benefits of data analysis for years, this is a new realm for Customs and trade. Armed with these insights, Customs operations management teams have the opportunity to increase their relevance, shifting their role from cost center to strategic player. Imagine what you could do if you had insights into potential duty savings, guarantee thresholds, stock levels, declaration status, and exactly where each of your shipments were in transit, all at the touch of a button.

Using machine learning to ensure data quality

Of course, the value of these insights is directly related to the quality of your data, and when you are gathering data from multiple sources, quality poses a major hurdle. You need to trust the parties providing the data as well as check and validate the data before it is fed into the centralized “single source of truth” system. Data should be verified to ensure it is accurate, relevant, complete, current, and consistent. This is accomplished most efficiently with machine learning. By gathering data from previous behaviors, machine learning builds statistical models based on that data and generates suggestions that become more and more targeted as more data is gathered.

While some may think of machine learning as a new technology, it was actually adopted by prominent industries back in the early 1990s. It is relied on by the medical sector to extrapolate accurate dosages for medication, by post offices to translate and process handwritten addresses, and by our friends Siri, Alexa, Cortana, and Google Assistant for speech recognition. Health organizations, scientists, and think tanks around the world turned to machine learning to fight the battle against COVID-19, using it to support forecasting models, contact tracing, and drug development.

In the realm of Customs, machine learning is an excellent tool to detect anomalies and mistakes in the data provided. It can be used to improve data quality, increase accuracy and efficiency in the Customs declarations process, reduce human error, detect significant differences in the value of goods, and suggest better classification codes to reduce costs – and the more it is used, the smarter it gets.

Machine learning use cases

Detect significant differences in declared Customs values of similar articles

The Customs value of a given article in a declaration is based on its invoice value plus or minus certain adjustable cost elements. The expected behavior is that the invoice value or adjustable cost elements may differ slightly due to changes in materials costs but will not fluctuate significantly. However, incorrect costs may be entered into the source system due to human error, resulting in substantial differences in value. These may get lost in the multitude of transactions, posing potential financial risks, including inflated duty and penalties for incorrect valuation. Machine learning can detect these anomalies by comparing the Customs value of each unique item identifier, or item ID, to the value of that ID in recent similar flows.

Detect differences in classification behavior for items

In master data, item IDs may have a classification code for a classification type, e.g., classification type TARIC with classification code 1511909900. The expected behavior is that different item IDs with very similar item descriptions would mostly have the same classification code for a given classification type. With tens of thousands of items, however, it is challenging for a classifier to detect small differences. Machine learning can help standardize the classification codes used for similar products and suggest better classification codes to reduce costs.

Prepare for changes with simulations

As new formalities take effect, knowing the implications in advance helps businesses prepare for necessary adjustments. Machine learning allows companies to simulate declarations and other procedures by applying proposed parameters. In the case of Brexit, machine learning can apply future changes to Customs flows, revealing areas that pose the risk of non-compliance and associated delays. Simulations are also helpful in preparing for free trade agreements, for example calculating and estimating origin savings if a preferential rate were to be claimed.

Centralized data plus automation

Adopting a software solution that houses all of their Customs data and allows them to automate Customs processes opens up a variety of opportunities to companies. When it comes to making a choice, companies naturally want to right-size the solution and pricing for their current needs but it is important to think globally to allow room for expansion. This means selecting a multi-country solution that serves as a centralized repository for all Customs data, allowing stakeholders from around the world role-based access in their native language. Solutions that take a modular approach allow companies to start with the features and functionality they most need, for example automating declarations, and add other capabilities, for example special procedures and analytics, as their capacities mature.

Customs automation can have a significant financial impact, providing a relatively rapid return on investment even for companies with limited trade. For example, a small-to-medium sized business handling an estimated 5,000 declarations a year across three different countries for direct filing can realize an annual savings of € 91,500 compared to using a Customs broker. Plus, Customs operations managers have the added benefit of regaining control of their data, opening even more doors for efficiency and growth. See the white paper Customs Brokers or Software Solution for further examples of cost savings.

What the future holds

Customs operations management teams are typically slow to adopt new technology; however, more and more are abandoning paper processes in favor of digital, catapulting the value of automation into the spotlight. Automation and machine learning provide efficiencies that are unmatched by manual processes and will therefore play a key role in the future management of Customs and trade processes, allowing all stakeholders involved to grow and operate more efficiently.

In addition, harnessing the value of data is quickly becoming essential to remain competitive. By adopting a Customs system that centralizes, standardizes, and consolidates data, companies and Customs operations management teams can regain control of their Customs operations as well as analyze data via dashboards and reports in order to improve Customs-related processes, promote efficiencies, and open the door for growth. All of these technological advances allow valuable resources to be allocated to more strategic initiatives, providing greater value to the business. Data-driven Customs solutions are already being used by market leaders. This is the future of Customs for all.

More information

[1] Enterprise resource planning (ERP) refers to a type of software that organizations use to standardize, streamline and integrate business processes across finance, human resources, procurement and other departments.

[2] A warehouse management system (WMS) is a software application that helps control and manage the day-to-day operations in a warehouse.