An interview with Sue Probert on the development and impact of semantic data models and their lasting impact on global trade.

In this interview, we talk to Sue Probert, who has just completed her second term as Chair of UN/CEFACT. UN/CEFACT (United Nations Centre for Trade Facilitation and Electronic Business) develops global standards and semantic data models to facilitate and harmonise international trade procedures and business processes. Over the past decades, Sue has not only witnessed but also shaped key developments in this field, playing a pivotal role in revolutionising how we exchange data globally. We look forward to exploring the insights and experiences from this life long journey.

You’ve had an extensive career in the field of electronic data exchange. Would you kindly start by telling us about your early career and how you got into this field?

Absolutely. I began my trade facilitation journey in the 1980s working for an IBM dealership in the UK. Back then, I was involved in developing a system that allowed exporters to create standardised export documents more efficiently using laser printing technology. This early experience really sparked my interest in the standardisation of data models and electronic trade. At that time, however, it was not yet a question of electronic data exchange between companies. Instead, the aim was to develop functions with which printers could efficiently print out the relevant business documents. So the documents were all still paper-based.

However, in the early 1990s, my company suddenly decided to stop developing the document creation system. I became redundant and found myself on the street, with no car, no laptop, no phone. It’s one of those times when you are forced to think about your future. Three months later, I had started a new business.

And I had negotiated with my previous employer that I could take over all the software that I had been responsible for developing together with most of the development team. So I started a tiny little company in my house. In the beginning, I had six employees, and when the kids went to school, their bedrooms were used as offices.

We focused on software solutions for international trade and through the UK SITPRO organisation, I joined a joint UN/ECE and OASIS ebXML project where I first encountered many people in the XML World. One of them wanted to work together with UN/CEFACT and develop new XML solutions for the international trade. Because of our expertise in both fields this resulted in one of the dotcom companies deciding to buy my company. That’s one of those really crucial events that resulted in a wonderful range of life experiences. I continued to work for this company for the next three years, partly in Silicon Valley and partly in the UK.

How did you then find your way to UN/CEFACT?

By selling the company, I was financially independent and could therefore decide freely about what I wanted to do next. The world of international trade continued to fascinate me. So I decided to give back some of my experience and started contributing directly to UN/CEFACT as an expert volunteer.

And you have come a long way since then. You recently finished your six-year chairmanship of UN/CEFACT, how do you look back on this time? Which of your contributions would you like to see have a longer-term impact?

Reference data models are definitely the most important thing I have worked on, not just in the last six years, but much longer. These Reference Data Models are structured in a meaningful way to represent data related to international supply chains. They ensure that the semantic data used in cross-border trade processes is well defined, standardised, and universally understood across different systems, organisations, and countries.

To understand this universal and standardised approach, let’s use the term “buyer” as an example. A buyer needs to be clearly defined so that everybody involved in a transaction knows very well who is responsible for the payment. The UN/CEFACT data model includes numerous attributes for the buyer party, many designed for general use, such as the company name, address and contact information. However, some attributes are only necessary in specific transactions, such as those involving regulated goods, special tax conditions, or unique contractual agreements. The semantic data models of UN/CEFACT are a kind of library in which all the important data relevant for international trade are defined in a standardised and comprehensive way.

Can you explain in more detail what advantages reference models have and how they are useful for companies in general?

A semantic reference data model allows trading partners to reuse the same data definitions regardless of the syntax format that they may be adopting for data exchange. This means that a company can switch from one syntax exchange format to another, or even adopt new formats in the future, without losing the underlying meaning of the data. This is particularly valuable for international trade, where you have to deal with different regulations and practices across country borders. Our models ensure that the semantic data definitions remain consistent and reliable, no matter where it’s used.

This reusability is the key advantage of semantic data models. In UN/CEAFCT we have continuously developed and expanded our international supply chain reference model and now offer a model that reflects processes in the international supply chain better than any other known supply chain model.

You’ve also been involved in the adoption of UN/EDIFACT, XML and JSON technologies. How did these change the landscape of data exchange?

On the one hand, each new syntax format certainly had a major impact on the technical implementation of data exchange. When XML became very popular around the 2000s and JSON a decade later, new standards and data formats were developed that were specifically tailored to the new data formats in terms of semantics and syntax.

On the other side, these changes have not fundamentally altered the operational processes within international trade. This continuity in processes highlights the importance for companies to shift their attention to semantic reference models that prioritise a clear understanding and alignment with these operational workflows.

So, would you say there’s a best data exchange format for companies to use today?

I wouldn’t say there’s only one best format. Each format—whether it’s UN/EDIFACT, XML, JSON, or even traditional paper forms—serves the same fundamental purpose: enabling data exchange between trading partners. The choice of exchange format often depends on the specific needs of the organisation and the technical expertise available. What’s important is that companies focus on the semantics of the data they are exchanging. If they get the semantics right, they can adapt to any format that comes along. But it is an important thing to remember that the developers in any organisation are often only experienced to implement what they’ve learned recently. And currently that’s most likely JSON and not semantic data models – this is a continual challenge in the real world. Another issue is that important lessons learned over the years are not always remembered over time.

What do you recommend companies do to ensure that they are well equipped for efficient data exchange?

My recommendation would be to focus on the semantics of your internal data systems and align them with international standards as much as possible. This alignment will make it much easier to exchange data with external partners, no matter what format is being used. If your internal systems are too rigid to change, then at least make sure that your external data exchanges are standardised.

When companies introduce a new ERP system or digitise processes, they all too often only think about their own internal procedures and lose sight of their external business partners. I find it amazing that they don’t think more about this. The question of how data is exchanged externally should be given a much higher priority.

And finally, what do you see as the biggest challenge for the future of data exchange?

The biggest challenge will be ensuring that all the different formats and technologies continue to be part of the picture. I spent my life meeting with people who think they’re doing something for the first time. They’re not. It’s a long, long journey and we all need to acknowledge both past and future in order to move forward. Otherwise we will just reinvent the same problems. There’s a lot of valuable data being exchanged in older formats like UN/EDIFACT, and we need to make sure that this remains accessible and usable. The future of data exchange needs to be inclusive of all relevant technologies.

Sue, thank you for the interview